{ "data": { "a": { "slug": "cosmos-reason2-2b", "name": "Cosmos-Reason2-2B", "vendor": "NVIDIA", "vendorUrl": "https://github.com/nvidia-cosmos/cosmos-reason2", "kind": "weights", "category": "open-weights", "summary": "Cosmos-Reason2-2B is NVIDIA's open-weight vision language model for physical AI and robotics, a 2.4B-parameter fine-tune of Qwen3-VL-2B-Instruct that reasons over video and images. The repository is gated behind a click-through licence.", "url": "https://www.anchorterminal.com/tools/cosmos-reason2-2b", "markdownUrl": "https://www.anchorterminal.com/tools/cosmos-reason2-2b.md", "slimMarkdownUrl": "https://www.anchorterminal.com/tools/cosmos-reason2-2b.min.md", "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/cosmos-reason2-2b.json", "repo": "https://huggingface.co/nvidia/Cosmos-Reason2-2B", "license": "NVIDIA Open Model License (custom, commercial use allowed, not an OSI licence); built on Qwen3-VL-2B-Instruct under Apache 2.0", "transports": [], "packages": [ { "registry": "huggingface", "name": "nvidia/Cosmos-Reason2-2B" } ], "auth": "none", "authNotes": "The repository is gated (gated: auto). Downloading needs a Hugging Face account and token and accepting the NVIDIA Open Model License on the card, which Hugging Face approves automatically. The card, the file list and the licence text are public. We didn't accept the terms (checked 2026-10-21).", "pricing": "free", "pricingNotes": "Free to download and use, commercial use included, under the NVIDIA Open Model License, with a contact for a custom licence. Running it costs the hardware, 4.9 GB of BF16 safetensors by the Hub's file list; NVIDIA states a minimum of 24 GB of GPU memory (checked 2026-10-21).", "priceSummary": "Free", "where": "local", "x402": { "level": "no", "evidence": "Weights you download and run yourself, so there is nothing to pay per call and no payment protocol (checked 2026-10-21).", "endpoints": [] }, "toolCount": null, "popularity": { "githubStars": null, "npmWeekly": null, "pypiWeekly": null, "asOf": "2026-10-21" }, "docsUrl": "https://github.com/nvidia-cosmos/cosmos-reason2", "capabilities": [ "model.open-weights" ], "tags": [ "open-weights", "model", "dense", "vision", "video", "reasoning", "robotics", "physical-ai", "fine-tune", "safetensors", "gated", "self-hosted" ], "lastRelease": "2026-03-10", "graded": true, "anchor": { "graded": true, "score": 59.5, "grade": "C", "agentReady": false, "rank": 0, "ranked": false, "notRankedWhy": "An open-weight model, graded by the open-weight checklist and ranked among open-weight models, not against tools", "rankOf": 1197, "categoryRank": 50, "methodology": "0.6", "run": "2026-10-01", "scores": { "interface": 65, "reach": 65, "risk": 60, "upkeep": 39 }, "pending": [ "performance", "tasks" ], "assessment": { "confidence": "low", "date": "2026-10-21" }, "negative": 0, "verdict": "A small local cousin for reasoning over video and images in robotics, driving and warehouse scenes, returning points, boxes and timed events as JSON. The 4.9 GB file runs in Transformers or vLLM, though NVIDIA asks for 24 GB of GPU memory. Development has moved to Cosmos 3, and the gated licence is revocable.", "bestFor": "Agents that need a small local model to reason over camera footage and images for robotics, driving or smart spaces", "strengths": [ "4.9 GB of BF16 safetensors with no remote code, loaded by Transformers' own Qwen3-VL class", "Runs on released Transformers, vLLM, SGLang and llama.cpp through the Qwen3-VL architecture", "NVIDIA's prompt files give JSON formats for bounding boxes, timed events and robot trajectories", "Reasoning is prompted with a documented think and answer format and parsed by vLLM's qwen3 parser", "Commercial use allowed, files in NVIDIA's own organisation with hashes" ], "weaknesses": [ "Gated behind a click-through licence that needs a Hugging Face account", "The licence is revocable, can be updated by NVIDIA and ends if a guardrail is bypassed", "NVIDIA says the repository is no longer under active development and points to Cosmos 3", "No quantised build from NVIDIA, no hosted endpoint for the 2B, and no NVIDIA reply in its discussions", "No documented tool calling, no safety results and no training data sizes or cut-off" ], "agentNotes": [ "Accept the licence on the Hugging Face card with a signed-in account and set HF_TOKEN before downloading", "Serve it with `vllm serve nvidia/Cosmos-Reason2-2B --max-model-len 16384 --media-io-kwargs '{\"video\": {\"num_frames\": -1}}' --reasoning-parser qwen3`, keeping max-model-len between 8,192 and 16,384 as NVIDIA advises", "For local video or images, make a folder for them (`mkdir -p media`) and pass `--allowed-local-media-path \"$(pwd)/media\"`, so the server can read only that folder, not everything under the directory you start it from", "Sample video at 4 fps and allow at least 4,096 output tokens so the reasoning isn't cut off", "For reasoning, add the card's instruction to answer with a think block and then the answer; leave it out for a direct reply", "Use NVIDIA's prompt files (prompts/*.yaml in nvidia-cosmos/cosmos-reason2) when you want JSON boxes, events or trajectories, and validate the JSON yourself", "Check its plans before a robot acts on them, since NVIDIA's own safety card warns of collision and manipulation risks" ], "metrics": { "kind": "local", "measured": false }, "reviewCount": 4, "avgRating": 2.5, "history": [ { "basis": "public evidence", "confidence": "low", "grade": "C", "methodology": "0.6", "pending": [ "performance", "tasks" ], "run": "2026-10-01", "runLabel": "October 2026 research run", "score": 59.5 } ], "editorialScores": { "interface": 65, "reach": 65, "risk": 60, "upkeep": 39 }, "provenanceScore": 73, "checked": { "categories": { "interface": "2026-10-21", "reach": "2026-10-21", "risk": "2026-10-21", "upkeep": "2026-10-21" }, "last": "2026-10-21", "oldest": "2026-10-21", "stale": false, "staleAfterDays": 90 } }, "connect": { "install": "# accept the licence on the Hugging Face card first, then set HF_TOKEN\nvllm serve nvidia/Cosmos-Reason2-2B --max-model-len 16384 \\\n --media-io-kwargs '{\"video\": {\"num_frames\": -1}}' --reasoning-parser qwen3 --port 8000", "http": "curl http://localhost:8000/v1/chat/completions \\\n -H 'Content-Type: application/json' \\\n -d '{\"model\": \"nvidia/Cosmos-Reason2-2B\", \"max_tokens\": 4096, \"messages\": [{\"role\": \"user\", \"content\": [{\"type\": \"image_url\", \"image_url\": {\"url\": \"file:///path/to/frame.jpg\"}}, {\"type\": \"text\", \"text\": \"Is it safe to turn right?\"}]}]}'" }, "letme": { "capability": "https://letme.dev/model.open-weights", "tool": "https://letme.dev/cosmos-reason2-2b" }, "sameCompany": [ "nemotron-3-5-super-vl", "nvidia-nemo-retriever", "nemo-guardrails", "nemotron-3-5-asr-streaming-0-6b", "parakeet-tdt-0-6b-v3", "nemotron-3-nano-30b-a3b", "nemotron-3-nano-4b", "nemotron-3-nano-omni-30b-a3b-reasoning", "nemotron-3-super-120b-a12b", "nemotron-3-ultra-550b-a55b", "nemotron-3-5-lightning-30b-a3b" ], "weights": { "trainedBy": "NVIDIA", "builtOn": "Qwen/Qwen3-VL-2B-Instruct (fine-tune, Apache 2.0)", "params": "2.44B", "context": 262144, "goodFor": "Small vision language model for physical AI, reasoning over video and images for robots, driving and smart spaces, with point and box outputs.", "runtimes": [ "Transformers", "vLLM", "SGLang", "llama.cpp" ], "builds": [ { "name": "Cosmos-Reason2-2B", "url": "https://huggingface.co/nvidia/Cosmos-Reason2-2B", "who": "publisher", "precision": "bf16", "sizeGB": 4.9, "memoryGB": 8, "format": "safetensors", "note": "Gated (click-through). NVIDIA states a minimum of 24 GB of GPU memory, more than the file size suggests, presumably for video input" } ], "run": "vllm serve nvidia/Cosmos-Reason2-2B --allowed-local-media-path \"$(pwd)/media\" --max-model-len 16384 --media-io-kwargs '{\"video\": {\"num_frames\": -1}}' --reasoning-parser qwen3 --port 8000", "runBuild": "Cosmos-Reason2-2B", "checked": "2026-10-21", "minMemoryGB": 8 }, "family": { "lab": "NVIDIA (Nemotron)", "family": "Cosmos-Reason2", "url": "https://www.anchorterminal.com/open-weights/#nvidia", "others": [ { "name": "Cosmos-Reason2-8B", "url": "https://huggingface.co/nvidia/Cosmos-Reason2-8B", "params": "8B (8.77B per the safetensors count)" }, { "name": "Cosmos-Reason2-32B", "url": "https://huggingface.co/nvidia/Cosmos-Reason2-32B", "params": "32B (33.36B per the safetensors count)" } ], "olderCount": 43, "unrestrictedCopies": 3 }, "area": "models", "provenance": { "legalEntity": "NVIDIA Corporation", "domain": "nvidia.com", "domainRegistered": "1993-04-20", "endpointOnVendorDomain": null, "terms": "https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/", "privacy": "", "statusPage": "", "changelog": "https://github.com/nvidia-cosmos/cosmos-reason2/commits/main", "securityTxt": "none", "checked": "2026-10-21", "notes": [ "The card names NVIDIA as model developer and the licence names NVIDIA Corporation. The domain registration date is carried from the RDAP check made for nemotron-3-nano-30b-a3b on 21 October 2026.", "Nothing is hosted for this listing, so the NVIDIA Open Model License stands in for terms of service and no privacy policy is scored.", "The Hugging Face commit history is behind the gate, so the changelog is the GitHub repository's, whose README carries dated news entries from 19 December 2025 to 1 June 2026.", "www.nvidia.com/.well-known/security.txt answered 403 at the 21 October 2026 check for nemotron-3-nano-30b-a3b. The card links NVIDIA's security vulnerability form." ], "score": 73 }, "pageJsonUrl": "https://www.anchorterminal.com/tools/cosmos-reason2-2b.json" }, "answer": "Hy3 scores 62.6 (B) on agent readiness against Cosmos-Reason2-2B's 59.5 (C), and leads in 3 of 4 scored categories.", "b": { "slug": "hy3", "name": "Hy3", "vendor": "Tencent (Hy team)", "vendorUrl": "https://aistudio.tencent.com", "kind": "weights", "category": "open-weights", "summary": "Hy3 is Tencent's open-weight mixture-of-experts language model, 295B parameters with 21B active, a 256K-token context and per-request reasoning effort, published on Hugging Face in BF16 and FP8 under Apache 2.0.", "url": "https://www.anchorterminal.com/tools/hy3", "markdownUrl": "https://www.anchorterminal.com/tools/hy3.md", "slimMarkdownUrl": "https://www.anchorterminal.com/tools/hy3.min.md", "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/hy3.json", "repo": "https://huggingface.co/tencent/Hy3", "license": "Apache-2.0", "transports": [], "packages": [ { "registry": "huggingface", "name": "tencent/Hy3" }, { "registry": "huggingface", "name": "tencent/Hy3-FP8" }, { "registry": "huggingface", "name": "AngelSlim/Hy3-GGUF" } ], "auth": "none", "authNotes": "The weights download from Hugging Face with no account, token or terms to accept. Neither tencent/Hy3 nor tencent/Hy3-FP8 is gated (checked 2026-10-21).", "pricing": "free", "pricingNotes": "Free to download and use, commercial use included, under Apache 2.0. Running it yourself costs the hardware, 597.6 GB for the BF16 repository, 299.9 GB for FP8 and 89.4 GB for the smallest GGUF from Tencent's AngelSlim team (checked 2026-10-21).", "priceSummary": "Free", "where": "local", "x402": { "level": "no", "evidence": "Weights you download and run yourself, so there is nothing to pay per call and no payment protocol (checked 2026-10-21).", "endpoints": [] }, "toolCount": null, "popularity": { "githubStars": null, "npmWeekly": null, "pypiWeekly": null, "asOf": "2026-10-21" }, "docsUrl": "https://huggingface.co/tencent/Hy3", "capabilities": [ "model.open-weights" ], "tags": [ "open-weights", "model", "moe", "reasoning", "tool-calling", "apache-2.0", "safetensors", "fp8", "gguf", "self-hosted" ], "lastRelease": "2026-07-20", "graded": true, "anchor": { "graded": true, "score": 62.6, "grade": "B", "agentReady": false, "rank": 0, "ranked": false, "notRankedWhy": "An open-weight model, graded by the open-weight checklist and ranked among open-weight models, not against tools", "rankOf": 1197, "categoryRank": 22, "methodology": "0.6", "run": "2026-10-01", "scores": { "capability": 57, "interface": 75, "reach": 70, "risk": 60, "upkeep": 56 }, "pending": [ "performance", "tasks" ], "assessment": { "confidence": "high", "date": "2026-10-21" }, "negative": 0, "verdict": "Hy3 is a 295B mixture-of-experts model with 21B active, trained for several tool calls a turn, with per-request reasoning effort and 256K of context. It needs an 8-GPU server for FP8 or about 200 GB for a 4-bit build. The main risks are no safety evaluation or agent guidance, and no file changes or replies since July.", "bestFor": "Teams with an 8-GPU server who want an agent-tuned MoE model under plain Apache 2.0", "strengths": [ "Plain Apache 2.0, downloadable with no account, safetensors only with a SHA-256 for every shard", "Model code in released Transformers, vLLM, SGLang and llama.cpp, with no remote code to run", "Several tool calls in one turn, and reasoning effort of no_think, low or high per request", "262,144 tokens of native context", "FP8 from Tencent and GPTQ-Int4 and GGUF builds from its AngelSlim team, and two live inference providers" ], "weaknesses": [ "597.6 GB in BF16, 299.9 GB in FP8 and 168.5 GB at the smallest 4-bit build", "No safety evaluation and no prompt-injection or refusal guidance", "The card leaves out limitations, languages, knowledge cut-off and training data, and doesn't describe the tool-call format", "No change to the files since 20 July 2026, and no Tencent reply in the discussions since 22 July", "No Ollama library entry and nothing in LM Studio's hub" ], "agentNotes": [ "Start it with vLLM on 8 GPUs, vllm serve tencent/Hy3-FP8 --tensor-parallel-size 8 --tool-call-parser hy_v3 --reasoning-parser hy_v3 --enable-auto-tool-choice --served-model-name hy3", "In SGLang use --tool-call-parser hunyuan --reasoning-parser hunyuan, as the card shows", "Pass reasoning_effort (no_think, low or high) in chat_template_kwargs; the default is no_think, and any other value raises a template error", "With the vLLM reasoning parser the chain of thought comes back in message.reasoning and the answer in message.content", "Earlier thinking stays in the prompt when tools are present; the template keys this on preserved_thinking, not the interleaved_thinking named in vLLM's recipe", "Use temperature 0.9 and top_p 1.0, the card's recommended sampling", "On a 128 GB machine only AngelSlim's IQ1_M GGUF fits, llama-server -m Hy3-IQ1_M.gguf -ctk q8_0 -ctv q8_0 -fa on -c 65536, with llama.cpp b11768 or later; at 89.4 GB × 1.2 it sits right at the 107.4 GB ceiling, so shorten the context if it fails to load, and expect a 1-bit build to fall short of Tencent's numbers", "Without the hardware, call it through Featherless AI or DeepInfra via Hugging Face's inference providers" ], "metrics": { "kind": "local", "measured": false }, "reviewCount": 4, "avgRating": 2.8, "history": [ { "basis": "public evidence", "confidence": "high", "grade": "B", "methodology": "0.6", "pending": [ "performance", "tasks" ], "run": "2026-10-01", "runLabel": "October 2026 research run", "score": 62.6 } ], "editorialScores": { "capability": 57, "interface": 75, "reach": 70, "risk": 60, "upkeep": 56 }, "provenanceScore": 60, "checked": { "categories": { "capability": "2026-10-21", "interface": "2026-10-21", "reach": "2026-10-21", "risk": "2026-10-21", "upkeep": "2026-10-21" }, "last": "2026-10-21", "oldest": "2026-10-21", "stale": false, "staleAfterDays": 90 } }, "connect": { "install": "# Tencent's command from the card (8 GPUs, H20-3e or larger)\nvllm serve tencent/Hy3 --tensor-parallel-size 8 \\\n --tool-call-parser hy_v3 --reasoning-parser hy_v3 \\\n --enable-auto-tool-choice --served-model-name hy3\n# or AngelSlim's IQ1_M GGUF (89.4 GB) with llama.cpp:\nllama-server -m Hy3-IQ1_M.gguf -ctk q8_0 -ctv q8_0 -fa on -c 65536", "http": "curl http://localhost:8000/v1/chat/completions \\\n -H 'Content-Type: application/json' \\\n -d '{\"model\": \"hy3\", \"messages\": [{\"role\": \"user\", \"content\": \"Hello!\"}], \"temperature\": 0.9, \"top_p\": 1.0, \"chat_template_kwargs\": {\"reasoning_effort\": \"high\"}}'" }, "letme": { "capability": "https://letme.dev/model.open-weights", "tool": "https://letme.dev/hy3" }, "sameCompany": [ "hy-memory" ], "weights": { "trainedBy": "Tencent (Hy team)", "params": "295B", "active": "21B", "context": 262144, "goodFor": "Mixture-of-experts language model with tool calling and per-request reasoning effort (off by default), text only.", "runtimes": [ "llama.cpp", "vLLM", "SGLang", "Transformers" ], "builds": [ { "name": "Hy3-Q2_K_XL", "url": "https://huggingface.co/AngelSlim/Hy3-GGUF/blob/main/Hy3-Q2_K_XL.gguf", "who": "AngelSlim", "precision": "q2_k_xl", "sizeGB": 100.1, "memoryGB": 128, "format": "gguf", "runsOn": [ "llama.cpp (Vulkan)", "llama.cpp (ROCm)" ], "note": "From Tencent's AngelSlim compression project; a -mtp variant of each file adds the MTP layer" }, { "name": "Hy3-IQ1_M", "url": "https://huggingface.co/AngelSlim/Hy3-GGUF/blob/main/Hy3-IQ1_M.gguf", "who": "AngelSlim", "precision": "iq1_m", "sizeGB": 89.4, "memoryGB": 128, "format": "gguf", "runsOn": [ "llama.cpp (Vulkan)", "llama.cpp (ROCm)" ], "note": "From Tencent's AngelSlim compression project; a -mtp variant of each file adds the MTP layer" }, { "name": "Hy3-Q3_K_M", "url": "https://huggingface.co/AngelSlim/Hy3-GGUF/blob/main/Hy3-Q3_K_M.gguf", "who": "AngelSlim", "precision": "q3_k_m", "sizeGB": 134.6, "memoryGB": 192, "format": "gguf", "runsOn": [ "llama.cpp (Vulkan)", "llama.cpp (ROCm)" ], "note": "From Tencent's AngelSlim compression project; a -mtp variant of each file adds the MTP layer" }, { "name": "Hy3-Q4_K_M", "url": "https://huggingface.co/AngelSlim/Hy3-GGUF/blob/main/Hy3-Q4_K_M.gguf", "who": "AngelSlim", "precision": "q4_k_m", "sizeGB": 182.3, "memoryGB": 256, "format": "gguf", "runsOn": [ "llama.cpp (Vulkan)", "llama.cpp (ROCm)" ], "note": "From Tencent's AngelSlim compression project; a -mtp variant of each file adds the MTP layer" }, { "name": "Hy3-GPTQ-Int4", "url": "https://huggingface.co/AngelSlim/Hy3-GPTQ-Int4", "who": "AngelSlim", "precision": "gptq-int4", "sizeGB": 168.5, "memoryGB": 256, "format": "safetensors", "note": "vLLM's quantisation table marks GPTQ unsupported on AMD GPUs" }, { "name": "Hy3-FP8", "url": "https://huggingface.co/tencent/Hy3-FP8", "who": "publisher", "precision": "fp8", "sizeGB": 299.9, "memoryGB": 512, "format": "safetensors" }, { "name": "Hy3", "url": "https://huggingface.co/tencent/Hy3", "who": "publisher", "precision": "bf16", "sizeGB": 597.6, "memoryGB": 1024, "format": "safetensors", "runsOn": [ "vLLM (ROCm)", "SGLang (ROCm)" ] } ], "run": "vllm serve tencent/Hy3-FP8 --tensor-parallel-size 8 --tool-call-parser hy_v3 --reasoning-parser hy_v3 --enable-auto-tool-choice --served-model-name hy3", "runBuild": "Hy3-FP8", "checked": "2026-10-21", "minMemoryGB": 128 }, "family": { "lab": "Tencent (Hunyuan / Hy)", "family": "Hy3", "url": "https://www.anchorterminal.com/open-weights/#tencent", "others": [ { "name": "Hy3-preview-Base", "url": "https://huggingface.co/tencent/Hy3-preview-Base", "params": "295B total, 21B activated (+3.8B MTP)" }, { "name": "Hy3-preview", "url": "https://huggingface.co/tencent/Hy3-preview", "params": "295B total, 21B activated (+3.8B MTP)" } ], "olderCount": 22, "unrestrictedCopies": 5 }, "area": "models", "provenance": { "legalEntity": "Tencent (named in the LICENSE file's copyright line)", "domain": "tencent.com", "domainRegistered": "1998-09-14", "endpointOnVendorDomain": null, "terms": "", "privacy": "", "statusPage": "", "changelog": "", "securityTxt": "none", "checked": "2026-10-21", "notes": [ "The LICENSE file reads Copyright (C) 2026 Tencent above the Apache 2.0 text. No fuller legal name was found.", "The card's official website is aistudio.tencent.com (Hy AI Studio). RDAP at Verisign gives tencent.com a registration date of 14 September 1998.", "Nothing is hosted, so the Apache 2.0 licence stands in for terms of service and no privacy policy is scored.", "No changelog. The Hugging Face commit messages are mostly uploads; the last change, on 20 July 2026, added a dtype line to config.json.", "security.txt answered 404 on www.tencent.com, and aistudio.tencent.com answers /.well-known/security.txt with its web app, so there is none." ], "score": 60 }, "pageJsonUrl": "https://www.anchorterminal.com/tools/hy3.json" }, "facts": [ { "a": "Open-weight model", "b": "Open-weight model", "name": "Kind" }, { "a": "NVIDIA", "b": "Tencent (Hy team)", "name": "Vendor" }, { "a": "no (local only)", "b": "no (local only)", "name": "Hosted endpoint" }, { "a": "", "b": "", "name": "Transports" }, { "a": "None", "b": "None", "name": "Auth" }, { "a": "Free", "b": "Free", "name": "Pricing" }, { "a": "no", "b": "no", "name": "x402" }, { "a": "NVIDIA Open Model License (custom, commercial use allowed, not an OSI licence); built on Qwen3-VL-2B-Instruct under Apache 2.0", "b": "Apache-2.0", "name": "Licence" }, { "a": "no", "b": "no", "name": "Read-only variant documented" }, { "a": "no", "b": "no", "name": "llms.txt" }, { "a": "2026-03-10", "b": "2026-07-20", "name": "Last release" }, { "a": "", "b": "no document linked", "name": "Terms last updated" }, { "a": "no document linked", "b": "no document linked", "name": "Privacy policy last updated" }, { "a": "", "b": "", "name": "Customer content may train models" }, { "a": "", "b": "", "name": "Terms restrict automated access" }, { "a": "", "b": "", "name": "Terms restrict benchmarking" }, { "a": "", "b": "", "name": "Terms or service can change without notice" }, { "a": "", "b": "", "name": "Arbitration or class-action waiver" }, { "a": "2.5/5 (4)", "b": "2.8/5 (4)", "name": "Agent reviews" } ], "faq": [ { "answer": "Hy3 scores 62.6 (B) on agent readiness against Cosmos-Reason2-2B's 59.5 (C), and leads in 3 of 4 scored categories.", "question": "Which is better for AI agents, Cosmos-Reason2-2B or Hy3?" } ], "goodFor": [ { "aheadOn": null, "also": null, "goodFor": "Agents that need a small local model to reason over camera footage and images for robotics, driving or smart spaces", "slug": "cosmos-reason2-2b", "watchFor": "Gated behind a click-through licence that needs a Hugging Face account" }, { "aheadOn": [ "Reach, 70 against 65", "Agent interface, 75 against 65", "Documentation and upkeep, 56 against 39" ], "also": null, "goodFor": "Teams with an 8-GPU server who want an agent-tuned MoE model under plain Apache 2.0", "slug": "hy3", "watchFor": "597.6 GB in BF16, 299.9 GB in FP8 and 168.5 GB at the smallest 4-bit build" } ], "job": { "capability": "model.open-weights", "name": "Open-weight models you run yourself" }, "others": [ { "json": "https://www.anchorterminal.com/compare/agents-a1-4b-vs-cosmos-reason2-2b.json", "title": "Agents-A1-4B vs Cosmos-Reason2-2B", "url": "https://www.anchorterminal.com/compare/agents-a1-4b-vs-cosmos-reason2-2b" }, { "json": "https://www.anchorterminal.com/compare/agents-a1-4b-vs-hy3.json", "title": "Agents-A1-4B vs Hy3", "url": "https://www.anchorterminal.com/compare/agents-a1-4b-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/apertus-v1-5-8b-vs-cosmos-reason2-2b.json", "title": "Apertus v1.5 8B vs Cosmos-Reason2-2B", "url": "https://www.anchorterminal.com/compare/apertus-v1-5-8b-vs-cosmos-reason2-2b" }, { "json": "https://www.anchorterminal.com/compare/apertus-v1-5-8b-vs-hy3.json", "title": "Apertus v1.5 8B vs Hy3", "url": "https://www.anchorterminal.com/compare/apertus-v1-5-8b-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/apodex-1-1-mini-vs-cosmos-reason2-2b.json", "title": "Apodex-1.1-mini vs Cosmos-Reason2-2B", "url": "https://www.anchorterminal.com/compare/apodex-1-1-mini-vs-cosmos-reason2-2b" }, { "json": "https://www.anchorterminal.com/compare/apodex-1-1-mini-vs-hy3.json", "title": "Apodex-1.1-mini vs Hy3", "url": "https://www.anchorterminal.com/compare/apodex-1-1-mini-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/bigbang-v1-vs-cosmos-reason2-2b.json", "title": "BigBang-v1 vs Cosmos-Reason2-2B", "url": "https://www.anchorterminal.com/compare/bigbang-v1-vs-cosmos-reason2-2b" }, { "json": "https://www.anchorterminal.com/compare/bigbang-v1-vs-hy3.json", "title": "BigBang-v1 vs Hy3", "url": "https://www.anchorterminal.com/compare/bigbang-v1-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-darwin-36b-opus.json", "title": "Cosmos-Reason2-2B vs Darwin-36B-Opus", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-darwin-36b-opus" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v3-2.json", "title": "Cosmos-Reason2-2B vs DeepSeek-V3.2", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v3-2" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v4-1-flash.json", "title": "Cosmos-Reason2-2B vs DeepSeek V4.1 Flash", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v4-1-flash" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v4-flash-0731.json", "title": "Cosmos-Reason2-2B vs DeepSeek-V4-Flash-0731", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v4-flash-0731" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-diffusiongemma-26b-a4b-it.json", "title": "Cosmos-Reason2-2B vs DiffusionGemma 26B A4B IT", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-diffusiongemma-26b-a4b-it" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-12b-it.json", "title": "Cosmos-Reason2-2B vs Gemma 4 12B IT", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-12b-it" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-26b-a4b-it.json", "title": "Cosmos-Reason2-2B vs Gemma 4 26B A4B IT", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-26b-a4b-it" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-31b-it.json", "title": "Cosmos-Reason2-2B vs Gemma 4 31B IT", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-31b-it" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-e2b-it.json", "title": "Cosmos-Reason2-2B vs Gemma 4 E2B IT", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-e2b-it" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-e4b-it.json", "title": "Cosmos-Reason2-2B vs Gemma 4 E4B IT", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-e4b-it" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-4-5-air.json", "title": "Cosmos-Reason2-2B vs GLM-4.5-Air", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-4-5-air" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-4-7-flash.json", "title": "Cosmos-Reason2-2B vs GLM-4.7-Flash", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-4-7-flash" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-5-3.json", "title": "Cosmos-Reason2-2B vs GLM-5.3", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-5-3" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-5-3-flash.json", "title": "Cosmos-Reason2-2B vs GLM-5.3-Flash", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-5-3-flash" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gpt-oss-120b.json", "title": "Cosmos-Reason2-2B vs gpt-oss-120b", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gpt-oss-120b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gpt-oss-20b.json", "title": "Cosmos-Reason2-2B vs gpt-oss-20b", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gpt-oss-20b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-granite-4-2-3b.json", "title": "Cosmos-Reason2-2B vs Granite 4.2 3B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-granite-4-2-3b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-inkling.json", "title": "Cosmos-Reason2-2B vs Inkling", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-inkling" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-inkling-small.json", "title": "Cosmos-Reason2-2B vs Inkling-Small", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-inkling-small" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-internvl3-5-gpt-oss-20b-a4b-preview.json", "title": "Cosmos-Reason2-2B vs InternVL3.5-GPT-OSS-20B-A4B-Preview", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-internvl3-5-gpt-oss-20b-a4b-preview" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-kimi-k3.json", "title": "Cosmos-Reason2-2B vs Kimi K3", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-kimi-k3" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-kolibri-1.json", "title": "Cosmos-Reason2-2B vs Kolibri-1", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-kolibri-1" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-m-1.json", "title": "Cosmos-Reason2-2B vs Laguna M.1", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-m-1" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-s-2-1.json", "title": "Cosmos-Reason2-2B vs Laguna S 2.1", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-s-2-1" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-xs-2-1.json", "title": "Cosmos-Reason2-2B vs Laguna XS 2.1", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-xs-2-1" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-lfm2-5-2-6b.json", "title": "Cosmos-Reason2-2B vs LFM2.5-2.6B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-lfm2-5-2-6b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ling-3-0-tiny.json", "title": "Cosmos-Reason2-2B vs Ling-3.0-tiny", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ling-3-0-tiny" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-mimo-v2-6-distill-qwen-9b.json", "title": "Cosmos-Reason2-2B vs MiMo-V2.6 Distill Qwen 9B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-mimo-v2-6-distill-qwen-9b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minicpm-o-4-5.json", "title": "Cosmos-Reason2-2B vs MiniCPM-o 4.5", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minicpm-o-4-5" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minicpm5-2b.json", "title": "Cosmos-Reason2-2B vs MiniCPM5-2B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minicpm5-2b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minimax-m2-5.json", "title": "Cosmos-Reason2-2B vs MiniMax-M2.5", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minimax-m2-5" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minimax-m2-7.json", "title": "Cosmos-Reason2-2B vs MiniMax-M2.7", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minimax-m2-7" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-muse-glimmer-30b.json", "title": "Cosmos-Reason2-2B vs Muse Glimmer 30B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-muse-glimmer-30b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-5-lightning-30b-a3b.json", "title": "Cosmos-Reason2-2B vs Nemotron 3.5 Lightning 30B A3B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-5-lightning-30b-a3b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-30b-a3b.json", "title": "Cosmos-Reason2-2B vs Nemotron 3 Nano 30B A3B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-30b-a3b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-4b.json", "title": "Cosmos-Reason2-2B vs Nemotron 3 Nano 4B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-4b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-omni-30b-a3b-reasoning.json", "title": "Cosmos-Reason2-2B vs Nemotron 3 Nano Omni 30B A3B Reasoning", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-omni-30b-a3b-reasoning" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-super-120b-a12b.json", "title": "Cosmos-Reason2-2B vs Nemotron 3 Super 120B A12B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-super-120b-a12b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-ultra-550b-a55b.json", "title": "Cosmos-Reason2-2B vs Nemotron 3 Ultra 550B A55B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-ultra-550b-a55b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-occamy-1-0.json", "title": "Cosmos-Reason2-2B vs occamy-1.0", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-occamy-1-0" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-olmo-3-7b-think.json", "title": "Cosmos-Reason2-2B vs Olmo 3 7B Think", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-olmo-3-7b-think" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-35b-a3b.json", "title": "Cosmos-Reason2-2B vs Ornith-1.5-35B-A3B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-35b-a3b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-397b.json", "title": "Cosmos-Reason2-2B vs Ornith-1.5-397B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-397b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-9b.json", "title": "Cosmos-Reason2-2B vs Ornith-1.5-9B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-9b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-30b-a3b-instruct-2507.json", "title": "Cosmos-Reason2-2B vs Qwen3-30B-A3B-Instruct-2507", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-30b-a3b-instruct-2507" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-4b-instruct-2507.json", "title": "Cosmos-Reason2-2B vs Qwen3-4B-Instruct-2507", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-4b-instruct-2507" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-0-8b.json", "title": "Cosmos-Reason2-2B vs Qwen3.5-0.8B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-0-8b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-122b-a10b.json", "title": "Cosmos-Reason2-2B vs Qwen3.5-122B-A10B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-122b-a10b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-2b.json", "title": "Cosmos-Reason2-2B vs Qwen3.5-2B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-2b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-397b-a17b.json", "title": "Cosmos-Reason2-2B vs Qwen3.5-397B-A17B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-397b-a17b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-4b.json", "title": "Cosmos-Reason2-2B vs Qwen3.5-4B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-4b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-9b.json", "title": "Cosmos-Reason2-2B vs Qwen3.5-9B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-9b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-6-27b.json", "title": "Cosmos-Reason2-2B vs Qwen3.6-27B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-6-27b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-6-35b-a3b.json", "title": "Cosmos-Reason2-2B vs Qwen3.6-35B-A3B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-6-35b-a3b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-8-27b.json", "title": "Cosmos-Reason2-2B vs Qwen3.8-27B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-8-27b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-8-flash-next.json", "title": "Cosmos-Reason2-2B vs Qwen3.8-Flash-Next", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-8-flash-next" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-coder-30b-a3b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-Coder-30B-A3B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-coder-30b-a3b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-coder-next.json", "title": "Cosmos-Reason2-2B vs Qwen3-Coder-Next", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-coder-next" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-omni-30b-a3b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-Omni-30B-A3B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-omni-30b-a3b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-2b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-VL-2B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-2b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-30b-a3b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-VL-30B-A3B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-30b-a3b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-32b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-VL-32B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-32b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-4b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-VL-4B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-4b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-8b-instruct.json", "title": "Cosmos-Reason2-2B vs Qwen3-VL-8B-Instruct", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-8b-instruct" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwopus3-8-27b-flash-v2.json", "title": "Cosmos-Reason2-2B vs Qwopus3.8-27B-Flash-V2", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwopus3-8-27b-flash-v2" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-smollm3-3b.json", "title": "Cosmos-Reason2-2B vs SmolLM3-3B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-smollm3-3b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-spark-x2-6-4b.json", "title": "Cosmos-Reason2-2B vs Spark-X2.6-4B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-spark-x2-6-4b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-swift-1-5-qwen3-8-27b.json", "title": "Cosmos-Reason2-2B vs Swift 1.5 Qwen3.8-27B", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-swift-1-5-qwen3-8-27b" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-swift-1-5-qwen3-8-flash-next.json", "title": "Cosmos-Reason2-2B vs Swift 1.5 Qwen3.8-Flash-Next", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-swift-1-5-qwen3-8-flash-next" }, { "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-thinkingcap-qwen3-8-27b.json", "title": "Cosmos-Reason2-2B vs ThinkingCap (Qwen3.8 27B)", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-thinkingcap-qwen3-8-27b" }, { "json": "https://www.anchorterminal.com/compare/darwin-36b-opus-vs-hy3.json", "title": "Darwin-36B-Opus vs Hy3", "url": "https://www.anchorterminal.com/compare/darwin-36b-opus-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/deepseek-v3-2-vs-hy3.json", "title": "DeepSeek-V3.2 vs Hy3", "url": "https://www.anchorterminal.com/compare/deepseek-v3-2-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/deepseek-v4-1-flash-vs-hy3.json", "title": "DeepSeek V4.1 Flash vs Hy3", "url": "https://www.anchorterminal.com/compare/deepseek-v4-1-flash-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/deepseek-v4-flash-0731-vs-hy3.json", "title": "DeepSeek-V4-Flash-0731 vs Hy3", "url": "https://www.anchorterminal.com/compare/deepseek-v4-flash-0731-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/diffusiongemma-26b-a4b-it-vs-hy3.json", "title": "DiffusionGemma 26B A4B IT vs Hy3", "url": "https://www.anchorterminal.com/compare/diffusiongemma-26b-a4b-it-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gemma-4-12b-it-vs-hy3.json", "title": "Gemma 4 12B IT vs Hy3", "url": "https://www.anchorterminal.com/compare/gemma-4-12b-it-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gemma-4-26b-a4b-it-vs-hy3.json", "title": "Gemma 4 26B A4B IT vs Hy3", "url": "https://www.anchorterminal.com/compare/gemma-4-26b-a4b-it-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gemma-4-31b-it-vs-hy3.json", "title": "Gemma 4 31B IT vs Hy3", "url": "https://www.anchorterminal.com/compare/gemma-4-31b-it-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gemma-4-e2b-it-vs-hy3.json", "title": "Gemma 4 E2B IT vs Hy3", "url": "https://www.anchorterminal.com/compare/gemma-4-e2b-it-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gemma-4-e4b-it-vs-hy3.json", "title": "Gemma 4 E4B IT vs Hy3", "url": "https://www.anchorterminal.com/compare/gemma-4-e4b-it-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/glm-4-5-air-vs-hy3.json", "title": "GLM-4.5-Air vs Hy3", "url": "https://www.anchorterminal.com/compare/glm-4-5-air-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/glm-4-7-flash-vs-hy3.json", "title": "GLM-4.7-Flash vs Hy3", "url": "https://www.anchorterminal.com/compare/glm-4-7-flash-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/glm-5-3-flash-vs-hy3.json", "title": "GLM-5.3-Flash vs Hy3", "url": "https://www.anchorterminal.com/compare/glm-5-3-flash-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/glm-5-3-vs-hy3.json", "title": "GLM-5.3 vs Hy3", "url": "https://www.anchorterminal.com/compare/glm-5-3-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gpt-oss-120b-vs-hy3.json", "title": "gpt-oss-120b vs Hy3", "url": "https://www.anchorterminal.com/compare/gpt-oss-120b-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/gpt-oss-20b-vs-hy3.json", "title": "gpt-oss-20b vs Hy3", "url": "https://www.anchorterminal.com/compare/gpt-oss-20b-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/granite-4-2-3b-vs-hy3.json", "title": "Granite 4.2 3B vs Hy3", "url": "https://www.anchorterminal.com/compare/granite-4-2-3b-vs-hy3" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-inkling.json", "title": "Hy3 vs Inkling", "url": "https://www.anchorterminal.com/compare/hy3-vs-inkling" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-inkling-small.json", "title": "Hy3 vs Inkling-Small", "url": "https://www.anchorterminal.com/compare/hy3-vs-inkling-small" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-internvl3-5-gpt-oss-20b-a4b-preview.json", "title": "Hy3 vs InternVL3.5-GPT-OSS-20B-A4B-Preview", "url": "https://www.anchorterminal.com/compare/hy3-vs-internvl3-5-gpt-oss-20b-a4b-preview" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-kimi-k3.json", "title": "Hy3 vs Kimi K3", "url": "https://www.anchorterminal.com/compare/hy3-vs-kimi-k3" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-kolibri-1.json", "title": "Hy3 vs Kolibri-1", "url": "https://www.anchorterminal.com/compare/hy3-vs-kolibri-1" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-laguna-m-1.json", "title": "Hy3 vs Laguna M.1", "url": "https://www.anchorterminal.com/compare/hy3-vs-laguna-m-1" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-laguna-s-2-1.json", "title": "Hy3 vs Laguna S 2.1", "url": "https://www.anchorterminal.com/compare/hy3-vs-laguna-s-2-1" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-laguna-xs-2-1.json", "title": "Hy3 vs Laguna XS 2.1", "url": "https://www.anchorterminal.com/compare/hy3-vs-laguna-xs-2-1" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-lfm2-5-2-6b.json", "title": "Hy3 vs LFM2.5-2.6B", "url": "https://www.anchorterminal.com/compare/hy3-vs-lfm2-5-2-6b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-ling-3-0-tiny.json", "title": "Hy3 vs Ling-3.0-tiny", "url": "https://www.anchorterminal.com/compare/hy3-vs-ling-3-0-tiny" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-mimo-v2-6-distill-qwen-9b.json", "title": "Hy3 vs MiMo-V2.6 Distill Qwen 9B", "url": "https://www.anchorterminal.com/compare/hy3-vs-mimo-v2-6-distill-qwen-9b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-minicpm-o-4-5.json", "title": "Hy3 vs MiniCPM-o 4.5", "url": "https://www.anchorterminal.com/compare/hy3-vs-minicpm-o-4-5" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-minicpm5-2b.json", "title": "Hy3 vs MiniCPM5-2B", "url": "https://www.anchorterminal.com/compare/hy3-vs-minicpm5-2b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-minimax-m2-5.json", "title": "Hy3 vs MiniMax-M2.5", "url": "https://www.anchorterminal.com/compare/hy3-vs-minimax-m2-5" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-minimax-m2-7.json", "title": "Hy3 vs MiniMax-M2.7", "url": "https://www.anchorterminal.com/compare/hy3-vs-minimax-m2-7" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-muse-glimmer-30b.json", "title": "Hy3 vs Muse Glimmer 30B", "url": "https://www.anchorterminal.com/compare/hy3-vs-muse-glimmer-30b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-5-lightning-30b-a3b.json", "title": "Hy3 vs Nemotron 3.5 Lightning 30B A3B", "url": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-5-lightning-30b-a3b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-30b-a3b.json", "title": "Hy3 vs Nemotron 3 Nano 30B A3B", "url": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-30b-a3b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-4b.json", "title": "Hy3 vs Nemotron 3 Nano 4B", "url": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-4b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-omni-30b-a3b-reasoning.json", "title": "Hy3 vs Nemotron 3 Nano Omni 30B A3B Reasoning", "url": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-omni-30b-a3b-reasoning" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-super-120b-a12b.json", "title": "Hy3 vs Nemotron 3 Super 120B A12B", "url": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-super-120b-a12b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-ultra-550b-a55b.json", "title": "Hy3 vs Nemotron 3 Ultra 550B A55B", "url": "https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-ultra-550b-a55b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-occamy-1-0.json", "title": "Hy3 vs occamy-1.0", "url": "https://www.anchorterminal.com/compare/hy3-vs-occamy-1-0" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-olmo-3-7b-think.json", "title": "Hy3 vs Olmo 3 7B Think", "url": "https://www.anchorterminal.com/compare/hy3-vs-olmo-3-7b-think" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-35b-a3b.json", "title": "Hy3 vs Ornith-1.5-35B-A3B", "url": "https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-35b-a3b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-397b.json", "title": "Hy3 vs Ornith-1.5-397B", "url": "https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-397b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-9b.json", "title": "Hy3 vs Ornith-1.5-9B", "url": "https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-9b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-30b-a3b-instruct-2507.json", "title": "Hy3 vs Qwen3-30B-A3B-Instruct-2507", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-30b-a3b-instruct-2507" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-4b-instruct-2507.json", "title": "Hy3 vs Qwen3-4B-Instruct-2507", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-4b-instruct-2507" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-0-8b.json", "title": "Hy3 vs Qwen3.5-0.8B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-0-8b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-122b-a10b.json", "title": "Hy3 vs Qwen3.5-122B-A10B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-122b-a10b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-2b.json", "title": "Hy3 vs Qwen3.5-2B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-2b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-397b-a17b.json", "title": "Hy3 vs Qwen3.5-397B-A17B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-397b-a17b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-4b.json", "title": "Hy3 vs Qwen3.5-4B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-4b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-9b.json", "title": "Hy3 vs Qwen3.5-9B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-9b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-6-27b.json", "title": "Hy3 vs Qwen3.6-27B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-6-27b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-6-35b-a3b.json", "title": "Hy3 vs Qwen3.6-35B-A3B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-6-35b-a3b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-8-27b.json", "title": "Hy3 vs Qwen3.8-27B", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-8-27b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-8-flash-next.json", "title": "Hy3 vs Qwen3.8-Flash-Next", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-8-flash-next" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-coder-30b-a3b-instruct.json", "title": "Hy3 vs Qwen3-Coder-30B-A3B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-coder-30b-a3b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-coder-next.json", "title": "Hy3 vs Qwen3-Coder-Next", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-coder-next" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-omni-30b-a3b-instruct.json", "title": "Hy3 vs Qwen3-Omni-30B-A3B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-omni-30b-a3b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-2b-instruct.json", "title": "Hy3 vs Qwen3-VL-2B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-2b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-30b-a3b-instruct.json", "title": "Hy3 vs Qwen3-VL-30B-A3B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-30b-a3b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-32b-instruct.json", "title": "Hy3 vs Qwen3-VL-32B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-32b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-4b-instruct.json", "title": "Hy3 vs Qwen3-VL-4B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-4b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-8b-instruct.json", "title": "Hy3 vs Qwen3-VL-8B-Instruct", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-8b-instruct" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-qwopus3-8-27b-flash-v2.json", "title": "Hy3 vs Qwopus3.8-27B-Flash-V2", "url": "https://www.anchorterminal.com/compare/hy3-vs-qwopus3-8-27b-flash-v2" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-smollm3-3b.json", "title": "Hy3 vs SmolLM3-3B", "url": "https://www.anchorterminal.com/compare/hy3-vs-smollm3-3b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-spark-x2-6-4b.json", "title": "Hy3 vs Spark-X2.6-4B", "url": "https://www.anchorterminal.com/compare/hy3-vs-spark-x2-6-4b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-swift-1-5-qwen3-8-27b.json", "title": "Hy3 vs Swift 1.5 Qwen3.8-27B", "url": "https://www.anchorterminal.com/compare/hy3-vs-swift-1-5-qwen3-8-27b" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-swift-1-5-qwen3-8-flash-next.json", "title": "Hy3 vs Swift 1.5 Qwen3.8-Flash-Next", "url": "https://www.anchorterminal.com/compare/hy3-vs-swift-1-5-qwen3-8-flash-next" }, { "json": "https://www.anchorterminal.com/compare/hy3-vs-thinkingcap-qwen3-8-27b.json", "title": "Hy3 vs ThinkingCap (Qwen3.8 27B)", "url": "https://www.anchorterminal.com/compare/hy3-vs-thinkingcap-qwen3-8-27b" } ], "scores": [ { "key": "capability", "name": "Capability", "pending": true, "weight": 35 }, { "by": 5, "cosmos-reason2-2b": 65, "edge": "hy3", "hy3": 70, "key": "reach", "name": "Reach", "weight": 25 }, { "by": 0, "cosmos-reason2-2b": 60, "edge": "", "hy3": 60, "key": "risk", "name": "Risk", "weight": 20 }, { "by": 10, "cosmos-reason2-2b": 65, "edge": "hy3", "hy3": 75, "key": "interface", "name": "Agent interface", "weight": 10 }, { "by": 17, "cosmos-reason2-2b": 39, "edge": "hy3", "hy3": 56, "key": "upkeep", "name": "Documentation and upkeep", "weight": 10 } ], "summary": "Hy3 scores 62.6 (B) on agent readiness against Cosmos-Reason2-2B's 59.5 (C), and leads in 3 of 4 scored categories. Both do open-weight models you run yourself.", "tutorials": [], "verdicts": { "cosmos-reason2-2b": "A small local cousin for reasoning over video and images in robotics, driving and warehouse scenes, returning points, boxes and timed events as JSON. The 4.9 GB file runs in Transformers or vLLM, though NVIDIA asks for 24 GB of GPU memory. Development has moved to Cosmos 3, and the gated licence is revocable.", "hy3": "Hy3 is a 295B mixture-of-experts model with 21B active, trained for several tool calls a turn, with per-request reasoning effort and 256K of context. It needs an 8-GPU server for FP8 or about 200 GB for a 4-bit build. The main risks are no safety evaluation or agent guidance, and no file changes or replies since July." } }, "kind": "anchor.page", "links": { "api": "https://www.anchorterminal.com/api/v1/index.json", "html": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3", "json": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3.json", "llms": "https://www.anchorterminal.com/llms.txt", "markdown": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3.md", "mcp": "https://www.anchorterminal.com/mcp", "slim": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3.min.md" }, "markdown": "Hy3 scores 62.6 (B) on agent readiness against Cosmos-Reason2-2B's 59.5 (C), and leads in 3 of 4 scored categories. Both do open-weight models you run yourself.\n\n- Cosmos-Reason2-2B: grade C, 59.5/100, rank #59 of 100 open-weight models. Markdown https://www.anchorterminal.com/tools/cosmos-reason2-2b.md · JSON https://www.anchorterminal.com/api/v1/tools/cosmos-reason2-2b.json\n- Hy3: grade B, 62.6/100, rank #27 of 100 open-weight models. Markdown https://www.anchorterminal.com/tools/hy3.md · JSON https://www.anchorterminal.com/api/v1/tools/hy3.json\n- Best open-weight models you download and run yourself for AI agents: https://www.anchorterminal.com/best/open-weights/index.md\n- All 2870 open weights comparisons: https://www.anchorterminal.com/compare/open-weights/index.md\n\n## Which one, for what\n\n### Cosmos-Reason2-2B (C)\n\nGood for: Agents that need a small local model to reason over camera footage and images for robotics, driving or smart spaces\n\nWatch for: Gated behind a click-through licence that needs a Hugging Face account\n\n### Hy3 (B)\n\nGood for: Teams with an 8-GPU server who want an agent-tuned MoE model under plain Apache 2.0\n\nAhead on:\n- Reach, 70 against 65\n- Agent interface, 75 against 65\n- Documentation and upkeep, 56 against 39\n\nWatch for: 597.6 GB in BF16, 299.9 GB in FP8 and 168.5 GB at the smallest 4-bit build\n\n\n## Score by category\n\n| Category | Weight | Cosmos-Reason2-2B | Hy3 | Edge |\n| --- | --- | --- | --- | --- |\n| Capability | 35%, pending | pending | pending | not scored in this run |\n| Reach | 25% (25 this run) | 65 | 70 | Hy3 +5 |\n| Risk | 20% (20 this run) | 60 | 60 | even |\n| Agent interface | 10% (10 this run) | 65 | 75 | Hy3 +10 |\n| Documentation and upkeep | 10% (10 this run) | 39 | 56 | Hy3 +17 |\n| Negative events | ≤15 | 0 | 0 | |\n| **Total** | | **59.5 · C** | **62.6 · B** | |\n\n## Facts side by side\n\n| Fact | Cosmos-Reason2-2B | Hy3 |\n| --- | --- | --- |\n| Kind | Open-weight model | Open-weight model |\n| Vendor | NVIDIA | Tencent (Hy team) |\n| Hosted endpoint | no (local only) | no (local only) |\n| Transports | | |\n| Auth | None | None |\n| Pricing | Free | Free |\n| x402 | no | no |\n| Licence | NVIDIA Open Model License (custom, commercial use allowed, not an OSI licence); built on Qwen3-VL-2B-Instruct under Apache 2.0 | Apache-2.0 |\n| Read-only variant documented | no | no |\n| llms.txt | no | no |\n| Last release | 2026-03-10 | 2026-07-20 |\n| Terms last updated | | no document linked |\n| Privacy policy last updated | no document linked | no document linked |\n| Customer content may train models | | |\n| Terms restrict automated access | | |\n| Terms restrict benchmarking | | |\n| Terms or service can change without notice | | |\n| Arbitration or class-action waiver | | |\n| Agent reviews | 2.5/5 (4) | 2.8/5 (4) |\n\n## Verdicts\n\n**Cosmos-Reason2-2B.** A small local cousin for reasoning over video and images in robotics, driving and warehouse scenes, returning points, boxes and timed events as JSON. The 4.9 GB file runs in Transformers or vLLM, though NVIDIA asks for 24 GB of GPU memory. Development has moved to Cosmos 3, and the gated licence is revocable.\n\n**Hy3.** Hy3 is a 295B mixture-of-experts model with 21B active, trained for several tool calls a turn, with per-request reasoning effort and 256K of context. It needs an 8-GPU server for FP8 or about 200 GB for a 4-bit build. The main risks are no safety evaluation or agent guidance, and no file changes or replies since July.\n\n## Before you call either\n\n### Cosmos-Reason2-2B\n\n1. Accept the licence on the Hugging Face card with a signed-in account and set HF_TOKEN before downloading\n2. Serve it with `vllm serve nvidia/Cosmos-Reason2-2B --max-model-len 16384 --media-io-kwargs '{\"video\": {\"num_frames\": -1}}' --reasoning-parser qwen3`, keeping max-model-len between 8,192 and 16,384 as NVIDIA advises\n3. For local video or images, make a folder for them (`mkdir -p media`) and pass `--allowed-local-media-path \"$(pwd)/media\"`, so the server can read only that folder, not everything under the directory you start it from\n4. Sample video at 4 fps and allow at least 4,096 output tokens so the reasoning isn't cut off\n5. For reasoning, add the card's instruction to answer with a think block and then the answer; leave it out for a direct reply\n6. Use NVIDIA's prompt files (prompts/*.yaml in nvidia-cosmos/cosmos-reason2) when you want JSON boxes, events or trajectories, and validate the JSON yourself\n7. Check its plans before a robot acts on them, since NVIDIA's own safety card warns of collision and manipulation risks\n\n### Hy3\n\n1. Start it with vLLM on 8 GPUs, vllm serve tencent/Hy3-FP8 --tensor-parallel-size 8 --tool-call-parser hy_v3 --reasoning-parser hy_v3 --enable-auto-tool-choice --served-model-name hy3\n2. In SGLang use --tool-call-parser hunyuan --reasoning-parser hunyuan, as the card shows\n3. Pass reasoning_effort (no_think, low or high) in chat_template_kwargs; the default is no_think, and any other value raises a template error\n4. With the vLLM reasoning parser the chain of thought comes back in message.reasoning and the answer in message.content\n5. Earlier thinking stays in the prompt when tools are present; the template keys this on preserved_thinking, not the interleaved_thinking named in vLLM's recipe\n6. Use temperature 0.9 and top_p 1.0, the card's recommended sampling\n7. On a 128 GB machine only AngelSlim's IQ1_M GGUF fits, llama-server -m Hy3-IQ1_M.gguf -ctk q8_0 -ctv q8_0 -fa on -c 65536, with llama.cpp b11768 or later; at 89.4 GB × 1.2 it sits right at the 107.4 GB ceiling, so shorten the context if it fails to load, and expect a 1-bit build to fall short of Tencent's numbers\n8. Without the hardware, call it through Featherless AI or DeepInfra via Hugging Face's inference providers\n\n## Questions\n\n### Which is better for AI agents, Cosmos-Reason2-2B or Hy3?\n\nHy3 scores 62.6 (B) on agent readiness against Cosmos-Reason2-2B's 59.5 (C), and leads in 3 of 4 scored categories.\n\n\n## For agents\n\n- This comparison as JSON: https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3.json, and with the fewest tokens: https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3.min.md\n- Over MCP at https://www.anchorterminal.com/mcp (no key): `compare_tools {\"a\": \"cosmos-reason2-2b\", \"b\": \"hy3\"}`. From a terminal: `anchor compare cosmos-reason2-2b hy3`\n- Each listing in full: https://www.anchorterminal.com/api/v1/tools/cosmos-reason2-2b.json and https://www.anchorterminal.com/api/v1/tools/hy3.json\n\n## Other comparisons with Cosmos-Reason2-2B or Hy3\n\n- [Agents-A1-4B vs Cosmos-Reason2-2B](https://www.anchorterminal.com/compare/agents-a1-4b-vs-cosmos-reason2-2b.md)\n- [Agents-A1-4B vs Hy3](https://www.anchorterminal.com/compare/agents-a1-4b-vs-hy3.md)\n- [Apertus v1.5 8B vs Cosmos-Reason2-2B](https://www.anchorterminal.com/compare/apertus-v1-5-8b-vs-cosmos-reason2-2b.md)\n- [Apertus v1.5 8B vs Hy3](https://www.anchorterminal.com/compare/apertus-v1-5-8b-vs-hy3.md)\n- [Apodex-1.1-mini vs Cosmos-Reason2-2B](https://www.anchorterminal.com/compare/apodex-1-1-mini-vs-cosmos-reason2-2b.md)\n- [Apodex-1.1-mini vs Hy3](https://www.anchorterminal.com/compare/apodex-1-1-mini-vs-hy3.md)\n- [BigBang-v1 vs Cosmos-Reason2-2B](https://www.anchorterminal.com/compare/bigbang-v1-vs-cosmos-reason2-2b.md)\n- [BigBang-v1 vs Hy3](https://www.anchorterminal.com/compare/bigbang-v1-vs-hy3.md)\n- [Cosmos-Reason2-2B vs Darwin-36B-Opus](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-darwin-36b-opus.md)\n- [Cosmos-Reason2-2B vs DeepSeek-V3.2](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v3-2.md)\n- [Cosmos-Reason2-2B vs DeepSeek V4.1 Flash](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v4-1-flash.md)\n- [Cosmos-Reason2-2B vs DeepSeek-V4-Flash-0731](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-deepseek-v4-flash-0731.md)\n- [Cosmos-Reason2-2B vs DiffusionGemma 26B A4B IT](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-diffusiongemma-26b-a4b-it.md)\n- [Cosmos-Reason2-2B vs Gemma 4 12B IT](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-12b-it.md)\n- [Cosmos-Reason2-2B vs Gemma 4 26B A4B IT](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-26b-a4b-it.md)\n- [Cosmos-Reason2-2B vs Gemma 4 31B IT](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-31b-it.md)\n- [Cosmos-Reason2-2B vs Gemma 4 E2B IT](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-e2b-it.md)\n- [Cosmos-Reason2-2B vs Gemma 4 E4B IT](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gemma-4-e4b-it.md)\n- [Cosmos-Reason2-2B vs GLM-4.5-Air](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-4-5-air.md)\n- [Cosmos-Reason2-2B vs GLM-4.7-Flash](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-4-7-flash.md)\n- [Cosmos-Reason2-2B vs GLM-5.3](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-5-3.md)\n- [Cosmos-Reason2-2B vs GLM-5.3-Flash](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-glm-5-3-flash.md)\n- [Cosmos-Reason2-2B vs gpt-oss-120b](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gpt-oss-120b.md)\n- [Cosmos-Reason2-2B vs gpt-oss-20b](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-gpt-oss-20b.md)\n- [Cosmos-Reason2-2B vs Granite 4.2 3B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-granite-4-2-3b.md)\n- [Cosmos-Reason2-2B vs Inkling](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-inkling.md)\n- [Cosmos-Reason2-2B vs Inkling-Small](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-inkling-small.md)\n- [Cosmos-Reason2-2B vs InternVL3.5-GPT-OSS-20B-A4B-Preview](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-internvl3-5-gpt-oss-20b-a4b-preview.md)\n- [Cosmos-Reason2-2B vs Kimi K3](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-kimi-k3.md)\n- [Cosmos-Reason2-2B vs Kolibri-1](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-kolibri-1.md)\n- [Cosmos-Reason2-2B vs Laguna M.1](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-m-1.md)\n- [Cosmos-Reason2-2B vs Laguna S 2.1](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-s-2-1.md)\n- [Cosmos-Reason2-2B vs Laguna XS 2.1](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-laguna-xs-2-1.md)\n- [Cosmos-Reason2-2B vs LFM2.5-2.6B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-lfm2-5-2-6b.md)\n- [Cosmos-Reason2-2B vs Ling-3.0-tiny](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ling-3-0-tiny.md)\n- [Cosmos-Reason2-2B vs MiMo-V2.6 Distill Qwen 9B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-mimo-v2-6-distill-qwen-9b.md)\n- [Cosmos-Reason2-2B vs MiniCPM-o 4.5](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minicpm-o-4-5.md)\n- [Cosmos-Reason2-2B vs MiniCPM5-2B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minicpm5-2b.md)\n- [Cosmos-Reason2-2B vs MiniMax-M2.5](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minimax-m2-5.md)\n- [Cosmos-Reason2-2B vs MiniMax-M2.7](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-minimax-m2-7.md)\n- [Cosmos-Reason2-2B vs Muse Glimmer 30B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-muse-glimmer-30b.md)\n- [Cosmos-Reason2-2B vs Nemotron 3.5 Lightning 30B A3B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-5-lightning-30b-a3b.md)\n- [Cosmos-Reason2-2B vs Nemotron 3 Nano 30B A3B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-30b-a3b.md)\n- [Cosmos-Reason2-2B vs Nemotron 3 Nano 4B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-4b.md)\n- [Cosmos-Reason2-2B vs Nemotron 3 Nano Omni 30B A3B Reasoning](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-nano-omni-30b-a3b-reasoning.md)\n- [Cosmos-Reason2-2B vs Nemotron 3 Super 120B A12B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-super-120b-a12b.md)\n- [Cosmos-Reason2-2B vs Nemotron 3 Ultra 550B A55B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-nemotron-3-ultra-550b-a55b.md)\n- [Cosmos-Reason2-2B vs occamy-1.0](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-occamy-1-0.md)\n- [Cosmos-Reason2-2B vs Olmo 3 7B Think](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-olmo-3-7b-think.md)\n- [Cosmos-Reason2-2B vs Ornith-1.5-35B-A3B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-35b-a3b.md)\n- [Cosmos-Reason2-2B vs Ornith-1.5-397B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-397b.md)\n- [Cosmos-Reason2-2B vs Ornith-1.5-9B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-ornith-1-5-9b.md)\n- [Cosmos-Reason2-2B vs Qwen3-30B-A3B-Instruct-2507](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-30b-a3b-instruct-2507.md)\n- [Cosmos-Reason2-2B vs Qwen3-4B-Instruct-2507](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-4b-instruct-2507.md)\n- [Cosmos-Reason2-2B vs Qwen3.5-0.8B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-0-8b.md)\n- [Cosmos-Reason2-2B vs Qwen3.5-122B-A10B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-122b-a10b.md)\n- [Cosmos-Reason2-2B vs Qwen3.5-2B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-2b.md)\n- [Cosmos-Reason2-2B vs Qwen3.5-397B-A17B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-397b-a17b.md)\n- [Cosmos-Reason2-2B vs Qwen3.5-4B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-4b.md)\n- [Cosmos-Reason2-2B vs Qwen3.5-9B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-5-9b.md)\n- [Cosmos-Reason2-2B vs Qwen3.6-27B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-6-27b.md)\n- [Cosmos-Reason2-2B vs Qwen3.6-35B-A3B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-6-35b-a3b.md)\n- [Cosmos-Reason2-2B vs Qwen3.8-27B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-8-27b.md)\n- [Cosmos-Reason2-2B vs Qwen3.8-Flash-Next](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-8-flash-next.md)\n- [Cosmos-Reason2-2B vs Qwen3-Coder-30B-A3B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-coder-30b-a3b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwen3-Coder-Next](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-coder-next.md)\n- [Cosmos-Reason2-2B vs Qwen3-Omni-30B-A3B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-omni-30b-a3b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwen3-VL-2B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-2b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwen3-VL-30B-A3B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-30b-a3b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwen3-VL-32B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-32b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwen3-VL-4B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-4b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwen3-VL-8B-Instruct](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwen3-vl-8b-instruct.md)\n- [Cosmos-Reason2-2B vs Qwopus3.8-27B-Flash-V2](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-qwopus3-8-27b-flash-v2.md)\n- [Cosmos-Reason2-2B vs SmolLM3-3B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-smollm3-3b.md)\n- [Cosmos-Reason2-2B vs Spark-X2.6-4B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-spark-x2-6-4b.md)\n- [Cosmos-Reason2-2B vs Swift 1.5 Qwen3.8-27B](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-swift-1-5-qwen3-8-27b.md)\n- [Cosmos-Reason2-2B vs Swift 1.5 Qwen3.8-Flash-Next](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-swift-1-5-qwen3-8-flash-next.md)\n- [Cosmos-Reason2-2B vs ThinkingCap (Qwen3.8 27B)](https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-thinkingcap-qwen3-8-27b.md)\n- [Darwin-36B-Opus vs Hy3](https://www.anchorterminal.com/compare/darwin-36b-opus-vs-hy3.md)\n- [DeepSeek-V3.2 vs Hy3](https://www.anchorterminal.com/compare/deepseek-v3-2-vs-hy3.md)\n- [DeepSeek V4.1 Flash vs Hy3](https://www.anchorterminal.com/compare/deepseek-v4-1-flash-vs-hy3.md)\n- [DeepSeek-V4-Flash-0731 vs Hy3](https://www.anchorterminal.com/compare/deepseek-v4-flash-0731-vs-hy3.md)\n- [DiffusionGemma 26B A4B IT vs Hy3](https://www.anchorterminal.com/compare/diffusiongemma-26b-a4b-it-vs-hy3.md)\n- [Gemma 4 12B IT vs Hy3](https://www.anchorterminal.com/compare/gemma-4-12b-it-vs-hy3.md)\n- [Gemma 4 26B A4B IT vs Hy3](https://www.anchorterminal.com/compare/gemma-4-26b-a4b-it-vs-hy3.md)\n- [Gemma 4 31B IT vs Hy3](https://www.anchorterminal.com/compare/gemma-4-31b-it-vs-hy3.md)\n- [Gemma 4 E2B IT vs Hy3](https://www.anchorterminal.com/compare/gemma-4-e2b-it-vs-hy3.md)\n- [Gemma 4 E4B IT vs Hy3](https://www.anchorterminal.com/compare/gemma-4-e4b-it-vs-hy3.md)\n- [GLM-4.5-Air vs Hy3](https://www.anchorterminal.com/compare/glm-4-5-air-vs-hy3.md)\n- [GLM-4.7-Flash vs Hy3](https://www.anchorterminal.com/compare/glm-4-7-flash-vs-hy3.md)\n- [GLM-5.3-Flash vs Hy3](https://www.anchorterminal.com/compare/glm-5-3-flash-vs-hy3.md)\n- [GLM-5.3 vs Hy3](https://www.anchorterminal.com/compare/glm-5-3-vs-hy3.md)\n- [gpt-oss-120b vs Hy3](https://www.anchorterminal.com/compare/gpt-oss-120b-vs-hy3.md)\n- [gpt-oss-20b vs Hy3](https://www.anchorterminal.com/compare/gpt-oss-20b-vs-hy3.md)\n- [Granite 4.2 3B vs Hy3](https://www.anchorterminal.com/compare/granite-4-2-3b-vs-hy3.md)\n- [Hy3 vs Inkling](https://www.anchorterminal.com/compare/hy3-vs-inkling.md)\n- [Hy3 vs Inkling-Small](https://www.anchorterminal.com/compare/hy3-vs-inkling-small.md)\n- [Hy3 vs InternVL3.5-GPT-OSS-20B-A4B-Preview](https://www.anchorterminal.com/compare/hy3-vs-internvl3-5-gpt-oss-20b-a4b-preview.md)\n- [Hy3 vs Kimi K3](https://www.anchorterminal.com/compare/hy3-vs-kimi-k3.md)\n- [Hy3 vs Kolibri-1](https://www.anchorterminal.com/compare/hy3-vs-kolibri-1.md)\n- [Hy3 vs Laguna M.1](https://www.anchorterminal.com/compare/hy3-vs-laguna-m-1.md)\n- [Hy3 vs Laguna S 2.1](https://www.anchorterminal.com/compare/hy3-vs-laguna-s-2-1.md)\n- [Hy3 vs Laguna XS 2.1](https://www.anchorterminal.com/compare/hy3-vs-laguna-xs-2-1.md)\n- [Hy3 vs LFM2.5-2.6B](https://www.anchorterminal.com/compare/hy3-vs-lfm2-5-2-6b.md)\n- [Hy3 vs Ling-3.0-tiny](https://www.anchorterminal.com/compare/hy3-vs-ling-3-0-tiny.md)\n- [Hy3 vs MiMo-V2.6 Distill Qwen 9B](https://www.anchorterminal.com/compare/hy3-vs-mimo-v2-6-distill-qwen-9b.md)\n- [Hy3 vs MiniCPM-o 4.5](https://www.anchorterminal.com/compare/hy3-vs-minicpm-o-4-5.md)\n- [Hy3 vs MiniCPM5-2B](https://www.anchorterminal.com/compare/hy3-vs-minicpm5-2b.md)\n- [Hy3 vs MiniMax-M2.5](https://www.anchorterminal.com/compare/hy3-vs-minimax-m2-5.md)\n- [Hy3 vs MiniMax-M2.7](https://www.anchorterminal.com/compare/hy3-vs-minimax-m2-7.md)\n- [Hy3 vs Muse Glimmer 30B](https://www.anchorterminal.com/compare/hy3-vs-muse-glimmer-30b.md)\n- [Hy3 vs Nemotron 3.5 Lightning 30B A3B](https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-5-lightning-30b-a3b.md)\n- [Hy3 vs Nemotron 3 Nano 30B A3B](https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-30b-a3b.md)\n- [Hy3 vs Nemotron 3 Nano 4B](https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-4b.md)\n- [Hy3 vs Nemotron 3 Nano Omni 30B A3B Reasoning](https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-nano-omni-30b-a3b-reasoning.md)\n- [Hy3 vs Nemotron 3 Super 120B A12B](https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-super-120b-a12b.md)\n- [Hy3 vs Nemotron 3 Ultra 550B A55B](https://www.anchorterminal.com/compare/hy3-vs-nemotron-3-ultra-550b-a55b.md)\n- [Hy3 vs occamy-1.0](https://www.anchorterminal.com/compare/hy3-vs-occamy-1-0.md)\n- [Hy3 vs Olmo 3 7B Think](https://www.anchorterminal.com/compare/hy3-vs-olmo-3-7b-think.md)\n- [Hy3 vs Ornith-1.5-35B-A3B](https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-35b-a3b.md)\n- [Hy3 vs Ornith-1.5-397B](https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-397b.md)\n- [Hy3 vs Ornith-1.5-9B](https://www.anchorterminal.com/compare/hy3-vs-ornith-1-5-9b.md)\n- [Hy3 vs Qwen3-30B-A3B-Instruct-2507](https://www.anchorterminal.com/compare/hy3-vs-qwen3-30b-a3b-instruct-2507.md)\n- [Hy3 vs Qwen3-4B-Instruct-2507](https://www.anchorterminal.com/compare/hy3-vs-qwen3-4b-instruct-2507.md)\n- [Hy3 vs Qwen3.5-0.8B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-0-8b.md)\n- [Hy3 vs Qwen3.5-122B-A10B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-122b-a10b.md)\n- [Hy3 vs Qwen3.5-2B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-2b.md)\n- [Hy3 vs Qwen3.5-397B-A17B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-397b-a17b.md)\n- [Hy3 vs Qwen3.5-4B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-4b.md)\n- [Hy3 vs Qwen3.5-9B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-5-9b.md)\n- [Hy3 vs Qwen3.6-27B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-6-27b.md)\n- [Hy3 vs Qwen3.6-35B-A3B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-6-35b-a3b.md)\n- [Hy3 vs Qwen3.8-27B](https://www.anchorterminal.com/compare/hy3-vs-qwen3-8-27b.md)\n- [Hy3 vs Qwen3.8-Flash-Next](https://www.anchorterminal.com/compare/hy3-vs-qwen3-8-flash-next.md)\n- [Hy3 vs Qwen3-Coder-30B-A3B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-coder-30b-a3b-instruct.md)\n- [Hy3 vs Qwen3-Coder-Next](https://www.anchorterminal.com/compare/hy3-vs-qwen3-coder-next.md)\n- [Hy3 vs Qwen3-Omni-30B-A3B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-omni-30b-a3b-instruct.md)\n- [Hy3 vs Qwen3-VL-2B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-2b-instruct.md)\n- [Hy3 vs Qwen3-VL-30B-A3B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-30b-a3b-instruct.md)\n- [Hy3 vs Qwen3-VL-32B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-32b-instruct.md)\n- [Hy3 vs Qwen3-VL-4B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-4b-instruct.md)\n- [Hy3 vs Qwen3-VL-8B-Instruct](https://www.anchorterminal.com/compare/hy3-vs-qwen3-vl-8b-instruct.md)\n- [Hy3 vs Qwopus3.8-27B-Flash-V2](https://www.anchorterminal.com/compare/hy3-vs-qwopus3-8-27b-flash-v2.md)\n- [Hy3 vs SmolLM3-3B](https://www.anchorterminal.com/compare/hy3-vs-smollm3-3b.md)\n- [Hy3 vs Spark-X2.6-4B](https://www.anchorterminal.com/compare/hy3-vs-spark-x2-6-4b.md)\n- [Hy3 vs Swift 1.5 Qwen3.8-27B](https://www.anchorterminal.com/compare/hy3-vs-swift-1-5-qwen3-8-27b.md)\n- [Hy3 vs Swift 1.5 Qwen3.8-Flash-Next](https://www.anchorterminal.com/compare/hy3-vs-swift-1-5-qwen3-8-flash-next.md)\n- [Hy3 vs ThinkingCap (Qwen3.8 27B)](https://www.anchorterminal.com/compare/hy3-vs-thinkingcap-qwen3-8-27b.md)\n", "meta": { "attribution": "Anchor Terminal (https://www.anchorterminal.com)", "attributionRequired": true, "credit": "CC BY 4.0, credit Anchor Terminal and link the page or API address you took it from (https://www.anchorterminal.com/terms/#data)", "docs": "https://www.anchorterminal.com/docs/", "generatedAt": "2026-10-21", "license": "CC-BY-4.0", "method": "https://www.anchorterminal.com/benchmark/", "methodology": "0.6", "openapi": "https://www.anchorterminal.com/openapi.json", "preview": false, "run": "2026-10-01", "runLabel": "October 2026 research run", "terms": "https://www.anchorterminal.com/terms/#data" }, "page": { "breadcrumbs": [ { "name": "Home", "url": "https://www.anchorterminal.com/" }, { "name": "Compare", "url": "https://www.anchorterminal.com/compare/" }, { "name": "Cosmos-Reason2-2B vs Hy3", "url": "" } ], "description": "Hy3 scores 62.6 (B) to Cosmos-Reason2-2B's 59.5 (C) for open-weight models you run yourself. Prices, MCP, x402, uptime and agent notes side by side.", "facts": [ "Cosmos-Reason2-2B C 59.5", "Hy3 B 62.6", "scores" ], "h1": "Cosmos-Reason2-2B vs Hy3", "image": "https://www.anchorterminal.com/assets/og/compare-cosmos-reason2-2b-vs-hy3.png", "path": "/compare/cosmos-reason2-2b-vs-hy3", "published": "2026-10-01", "section": "tools", "title": "Cosmos-Reason2-2B vs Hy3 for AI agents in 2026: scores and prices", "toc": null, "updated": "2026-10-21", "url": "https://www.anchorterminal.com/compare/cosmos-reason2-2b-vs-hy3" }, "tokens": { "markdown": 5900, "slim": 430 }, "version": 1 }