{
  "listings": [
    {
      "slug": "gemma-4-12b-it",
      "name": "Gemma 4 12B IT",
      "company": "Google LLC (RDAP registrant of gemma.ai as recorded for the first batch's Gemma 4 listings; the card names Google DeepMind as author)",
      "grade": "BB",
      "fit": "yes",
      "why": "Thinking is off by default, the Google Q4_0 GGUF is 7 GB, and Ollama's gemma4:12b (7.7 GB) carries the tools and thinking badges, so it sits beside a main model.",
      "source": "https://huggingface.co/google/gemma-4-12B-it/raw/main/chat_template.jinja",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gemma-4-12b-it"
    },
    {
      "slug": "gemma-4-e2b-it",
      "name": "Gemma 4 E2B IT",
      "company": "Google LLC (RDAP registrant of gemma.ai as recorded for the first batch's Gemma 4 listings; the card names Google DeepMind as author)",
      "grade": "C",
      "fit": "yes",
      "why": "3.35 GB Q4_0 from Google, thinking off by default, Ollama gemma4:e2b (4.6 GB) with tools and thinking badges; built for precisely this spare-memory role.",
      "source": "https://huggingface.co/api/models/google/gemma-4-E2B-it-qat-q4_0-gguf?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gemma-4-e2b-it"
    },
    {
      "slug": "gemma-4-e4b-it",
      "name": "Gemma 4 E4B IT",
      "company": "Google LLC (RDAP registrant of gemma.ai as recorded for the first batch's Gemma 4 listings; the card names Google DeepMind as author)",
      "grade": "B",
      "fit": "yes",
      "why": "5.15 GB Q4_0 from Google, thinking off by default, Ollama gemma4:e4b (6.6 GB) with tools and thinking badges, and acc-45's agent task passed.",
      "source": "https://huggingface.co/api/models/google/gemma-4-E4B-it-qat-q4_0-gguf?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gemma-4-e4b-it"
    },
    {
      "slug": "granite-4-2-3b",
      "name": "Granite 4.2 3B",
      "company": "International Business Machines Corporation (per IBM's naming guidance, as read for granite-guardian)",
      "grade": "C",
      "fit": "yes",
      "why": "acc-26 found Granite 4 routed 33 voice prompts at 100 percent in 0.54 seconds on a 16 GB GPU; IBM ships GGUFs from 1.5 GB, Ollama has granite4.2:3b (2.2 GB) and enable_thinking false is documented.",
      "source": "https://bitworking.org/news/2026/05/surprising-things-i-learned-putting-together-a-home-brain/",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/granite-4-2-3b"
    },
    {
      "slug": "ling-3-0-tiny",
      "name": "Ling-3.0-tiny",
      "company": "Ant Intelligence (Hang Zhou) Technology Co., Ltd (named as provider in the training-content summary); inclusionAI, Ant Group in the site footer of inclusion-ai.org",
      "grade": "C",
      "fit": "yes",
      "why": "4.8 GB Q4_K_M co-resides easily, enable_thinking false switches reasoning off per request, and an account measured 40 tokens per second on a CPU and over 200 on an RTX 4070.",
      "source": "https://huggingface.co/api/models/inclusionAI/Ling-3.0-tiny-GGUF/tree/main?recursive=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/ling-3-0-tiny"
    },
    {
      "slug": "nemotron-3-5-bolt-e4b-4gb",
      "name": "NVIDIA Nemotron 3.5 Bolt E4B 4GB",
      "company": "NVIDIA Corporation",
      "grade": "C",
      "fit": "yes",
      "why": "6.2 GB Q4_K_M from NVIDIA, enable_thinking false per request, a Brevity Benchmark row and an Ollama e4b tag of 7.2 GB to 7.9 GB with the tools badge.",
      "source": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-4GB/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-5-bolt-e4b-4gb"
    },
    {
      "slug": "nemotron-3-nano-4b",
      "name": "Nemotron 3 Nano 4B",
      "company": "NVIDIA Corporation",
      "grade": "C",
      "fit": "yes",
      "why": "2.8 GB Ollama tag and NVIDIA Q4_K_M GGUF, enable_thinking false per request, so it sits beside a main model in under 3 GB; English only.",
      "source": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-nano-4b"
    },
    {
      "slug": "qwen3-4b-instruct-2507",
      "name": "Qwen3-4B-Instruct-2507",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "C",
      "fit": "yes",
      "why": "Never thinks, 2.5 GB Ollama 4b-instruct tag with the tools badge, 262K context; a routing or summary call needs no switch.",
      "source": "https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-4b-instruct-2507"
    },
    {
      "slug": "qwen3-5-0-8b",
      "name": "Qwen3.5-0.8B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "D",
      "fit": "yes",
      "why": "Thinking is off by default; the card warns this size falls into thinking loops if you turn it on. Ollama qwen3.5:0.8b is 1.0 GB with the tools badge, so it sits beside a main model.",
      "source": "https://huggingface.co/Qwen/Qwen3.5-0.8B",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-5-0-8b"
    },
    {
      "slug": "qwen3-5-2b",
      "name": "Qwen3.5-2B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "D",
      "fit": "yes",
      "why": "Thinking is off by default; the card warns this size falls into thinking loops if you turn it on. Ollama qwen3.5:2b is 2.7 GB with the tools badge, so it sits beside a main model.",
      "source": "https://huggingface.co/Qwen/Qwen3.5-2B",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-5-2b"
    },
    {
      "slug": "qwen3-5-4b",
      "name": "Qwen3.5-4B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "C",
      "fit": "yes",
      "why": "Thinking switched off per request with enable_thinking false, Ollama qwen3.5:4b at about 3 GB with the tools badge, 2 GB of KV at 64K.",
      "source": "https://huggingface.co/Qwen/Qwen3.5-4B",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-5-4b"
    },
    {
      "slug": "qwen3-coder-30b-a3b-instruct",
      "name": "Qwen3-Coder-30B-A3B-Instruct",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "C",
      "fit": "yes",
      "why": "No thinking mode, 19 GB Ollama tag with the tools badge; acc-51 kept it resident beside a 32B planner at about 40 GB total.",
      "source": "https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-coder-30b-a3b-instruct"
    },
    {
      "slug": "qwen3-vl-2b-instruct",
      "name": "Qwen3-VL-2B-Instruct",
      "company": "Alibaba Cloud (per the first batch's RDAP check of qwen.ai; this model's files name no copyright holder)",
      "grade": "C",
      "fit": "yes",
      "why": "No thinking mode, publisher Q4_K_M GGUF of 1.1 GB, Ollama qwen3-vl:2b-instruct at 1.9 GB with the tools badge.",
      "source": "https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-vl-2b-instruct"
    },
    {
      "slug": "qwen3-vl-4b-instruct",
      "name": "Qwen3-VL-4B-Instruct",
      "company": "Alibaba Cloud (per the first batch's RDAP check of qwen.ai; this model's files name no copyright holder)",
      "grade": "D",
      "fit": "yes",
      "why": "No thinking mode, publisher Q4_K_M GGUF of 2.5 GB, Ollama qwen3-vl:4b-instruct at 3.3 GB with the tools badge.",
      "source": "https://huggingface.co/Qwen/Qwen3-VL-4B-Instruct",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-vl-4b-instruct"
    },
    {
      "slug": "qwen3-vl-8b-instruct",
      "name": "Qwen3-VL-8B-Instruct",
      "company": "Alibaba Cloud (per the pilot's RDAP check of qwen.ai; this model's files name no copyright holder)",
      "grade": "C",
      "fit": "yes",
      "why": "No thinking mode, publisher Q4_K_M GGUF of 5.0 GB, Ollama qwen3-vl:8b-instruct at 6.2 GB with the tools badge.",
      "source": "https://huggingface.co/Qwen/Qwen3-VL-8B-Instruct",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-vl-8b-instruct"
    },
    {
      "slug": "spark-x2-6-4b",
      "name": "Spark-X2.6-4B",
      "company": "XHToken (the licence's copyright line); the Hugging Face organisation's name is SparkLLM",
      "grade": "C",
      "fit": "yes",
      "why": "enable_thinking false or reasoning_effort low per request, publisher Q4_K_M GGUF of 3.5 GB, Ollama SparkLLM/Spark-X2.6-4B entry.",
      "source": "https://huggingface.co/XHToken/Spark-X2.6-4B",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/spark-x2-6-4b"
    },
    {
      "slug": "agents-a1-4b",
      "name": "Agents-A1-4B",
      "company": "Shanghai Artificial Intelligence Laboratory (the InternAgent team there, per the card's recommended system prompt and the template's default prompt); the GitHub LICENSE leaves the copyright line as the Apache template",
      "grade": "C",
      "fit": "caveat",
      "why": "2.7 GB Q4_K_M build from the publisher and enable_thinking false in the template, but thinking is on by default and the card never documents the switch; no Ollama entry.",
      "source": "https://huggingface.co/api/models/InternScience/Agents-A1-4B-Q4_K_M-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/agents-a1-4b"
    },
    {
      "slug": "apodex-1-1-mini",
      "name": "Apodex-1.1-mini",
      "company": "Apodex (the site's copyright line reads 2026 Apodex)",
      "grade": "C",
      "fit": "caveat",
      "why": "A 21.7 GB Q4_K_M GGUF fits 25 GB of spare memory and enable_thinking false is in the template, but thinking is on by default, the card does not document the switch and there is no Ollama entry.",
      "source": "https://huggingface.co/api/models/apodex/Apodex-1.1-mini-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/apodex-1-1-mini"
    },
    {
      "slug": "bigbang-v1",
      "name": "BigBang-v1",
      "company": "The Endless Frontier (endless-frontier)",
      "grade": "C",
      "fit": "caveat",
      "why": "No publisher GGUF; bartowski's Q4_K_M is 21.9 GB and the template's enable_thinking switch is undocumented on the card. Thinking defaults on.",
      "source": "https://huggingface.co/api/models/bartowski/endless-frontier_BigBang-v1-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/bigbang-v1"
    },
    {
      "slug": "darwin-36b-opus",
      "name": "Darwin-36B-Opus",
      "company": "VIDRAFT Inc. (named on vidraft.net); the card's citation names FINAL-Bench and VIDRAFT_LAB",
      "grade": "C",
      "fit": "caveat",
      "why": "Publisher GGUFs down to IQ1_M (8.2 GB) exist and enable_thinking false is in the template, but the card presents it as an always-thinking reasoning model and never names the switch.",
      "source": "https://huggingface.co/api/models/FINAL-Bench/POCKET-35B-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/darwin-36b-opus"
    },
    {
      "slug": "gemma-4-26b-a4b-it",
      "name": "Gemma 4 26B A4B IT",
      "company": "Google LLC (RDAP registrant of gemma.ai; the card names Google DeepMind as author)",
      "grade": "BB",
      "fit": "caveat",
      "why": "Thinking off by default and an Ollama tag with tools, but the Q4_0 build is 14.4 GB (16 GB on Ollama), so it needs the top of the 2 to 25 GB spare band.",
      "source": "https://huggingface.co/api/models/google/gemma-4-26B-A4B-it-qat-q4_0-gguf?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gemma-4-26b-a4b-it"
    },
    {
      "slug": "gemma-4-31b-it",
      "name": "Gemma 4 31B IT",
      "company": "Google LLC (the registrant of gemma.ai; the card names Google DeepMind as author)",
      "grade": "BB",
      "fit": "caveat",
      "why": "Thinking off by default and an Ollama tag with tools, but 17.6 to 19 GB at Q4 uses most of a 25 GB spare band; a 31B dense is slow for a 3 second routing reply.",
      "source": "https://huggingface.co/api/models/google/gemma-4-31B-it-qat-q4_0-gguf?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gemma-4-31b-it"
    },
    {
      "slug": "glm-4-7-flash",
      "name": "GLM-4.7-Flash",
      "company": "JINGSHENG HENGXING TECHNOLOGY PTE.LTD (operator of Z.ai per its terms of use, and registrant of z.ai)",
      "grade": "B",
      "fit": "caveat",
      "why": "enable_thinking false is in the template and Unsloth says 24 GB of memory at 4-bit, but 17.5 to 19 GB of weights takes most of a 25 GB spare band.",
      "source": "https://huggingface.co/zai-org/GLM-4.7-Flash/raw/main/chat_template.jinja",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/glm-4-7-flash"
    },
    {
      "slug": "gpt-oss-20b",
      "name": "gpt-oss-20b",
      "company": "OpenAI (named in USAGE_POLICY and the model card; the LICENSE file's copyright line is the unfilled Apache template)",
      "grade": "B",
      "fit": "caveat",
      "why": "12 to 14 GB fits a 25 GB spare band and reasoning low is cheap, but thinking cannot be switched off, so a routing reply still carries an analysis channel.",
      "source": "https://huggingface.co/openai/gpt-oss-20b/raw/main/chat_template.jinja",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gpt-oss-20b"
    },
    {
      "slug": "laguna-xs-2-1",
      "name": "Laguna XS 2.1",
      "company": "Poolside, Inc. (named in Poolside's Terms of Use)",
      "grade": "C",
      "fit": "caveat",
      "why": "Thinking is off by default and 3B are active, but the smallest build is 20 GB, at the top of the spare-memory range.",
      "source": "https://huggingface.co/api/models/poolside/Laguna-XS-2.1-GGUF/tree/main?recursive=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/laguna-xs-2-1"
    },
    {
      "slug": "lfm2-5-2-6b",
      "name": "LFM2.5-2.6B",
      "company": "Liquid AI, Inc. (the licensor named in the LFM Open License)",
      "grade": "D",
      "fit": "caveat",
      "why": "1.6 GB at Q4_0 and an Ollama entry with tools, but the card says it always thinks and the template forces a think tag, so routing calls carry reasoning time.",
      "source": "https://huggingface.co/LiquidAI/LFM2.5-2.6B/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/lfm2-5-2-6b"
    },
    {
      "slug": "mimo-v2-6-distill-qwen-9b",
      "name": "MiMo-V2.6 Distill Qwen 9B",
      "company": "Xiaomi (the card and report name Xiaomi MiMo; no LICENSE file names a copyright holder)",
      "grade": "C",
      "fit": "caveat",
      "why": "6.6 GB and enable_thinking false empties the think block, but it is an SFT checkpoint meant for RL and non-stop generation is reported.",
      "source": "https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B/raw/main/chat_template.jinja",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/mimo-v2-6-distill-qwen-9b"
    },
    {
      "slug": "minicpm-o-4-5",
      "name": "MiniCPM-o 4.5",
      "company": "OpenBMB (named as the MiniCPM-o team on the technical report; the GitHub LICENSE is the unfilled Apache template)",
      "grade": "C",
      "fit": "caveat",
      "why": "Q4_0 is 4.8 GB and the template empties the think block on enable_thinking false, but tool calling is undocumented and the Ollama build carries only a vision badge.",
      "source": "https://huggingface.co/api/models/openbmb/MiniCPM-o-4_5-gguf/tree/main?recursive=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/minicpm-o-4-5"
    },
    {
      "slug": "minicpm5-2b",
      "name": "MiniCPM5-2B",
      "company": "OpenBMB (named in the GitHub LICENSE's copyright line)",
      "grade": "C",
      "fit": "caveat",
      "why": "1.56 GB Q4_K_M and an Ollama entry, but every answer opens with reasoning; the template accepts enable_thinking false while the card documents only Think mode, and discussions report loops.",
      "source": "https://huggingface.co/api/models/openbmb/MiniCPM5-2B-GGUF/tree/main?recursive=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/minicpm5-2b"
    },
    {
      "slug": "muse-glimmer-30b",
      "name": "Muse Glimmer 30B",
      "company": "Meta (the verified Hugging Face organisation is \"Meta Inc.\", the card names Meta Superintelligence Lab, and the licence carries no copyright line)",
      "grade": "BB",
      "fit": "caveat",
      "why": "The 17 GB build fits a 25 GB spare budget, but reasoning_strength bottoms out at low with no off value, so a routing call still thinks.",
      "source": "https://dev.meta.ai/docs/muse-glimmer/prompting.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/muse-glimmer-30b"
    },
    {
      "slug": "nanbeige-4-3-33b",
      "name": "Nanbeige4.3-33B",
      "company": "BOSS Zhipin (Kanzhun Limited), unconfirmed",
      "grade": "C",
      "fit": "caveat",
      "why": "enable_thinking false turns reasoning off, but the smallest usable 4-bit build is about 19 GB, most of a 25 GB spare budget.",
      "source": "https://huggingface.co/Nanbeige/Nanbeige4.3-33B/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nanbeige-4-3-33b"
    },
    {
      "slug": "nemotron-3-5-lightning-30b-a3b",
      "name": "Nemotron 3.5 Lightning 30B A3B",
      "company": "NVIDIA Corporation",
      "grade": "B",
      "fit": "caveat",
      "why": "enable_thinking false turns reasoning off and 3B active is fast, but the smallest build is 18.9 GB, most of a 25 GB spare budget.",
      "source": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-5-lightning-30b-a3b"
    },
    {
      "slug": "nemotron-3-nano-30b-a3b",
      "name": "Nemotron 3 Nano 30B A3B",
      "company": "NVIDIA Corporation",
      "grade": "B",
      "fit": "caveat",
      "why": "enable_thinking false and a reasoning_budget example on the card, 3.5B active, but 18 GB is most of a 25 GB spare budget.",
      "source": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-nano-30b-a3b"
    },
    {
      "slug": "nemotron-3-nano-omni-30b-a3b-reasoning",
      "name": "Nemotron 3 Nano Omni 30B A3B Reasoning",
      "company": "NVIDIA Corporation",
      "grade": "C",
      "fit": "caveat",
      "why": "enable_thinking false gives an instruct mode, but the smallest GGUF is 18.5 GB, most of a 25 GB spare budget.",
      "source": "https://huggingface.co/nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-nano-omni-30b-a3b-reasoning"
    },
    {
      "slug": "occamy-1-0",
      "name": "occamy-1.0",
      "company": "Accio (Accio-Lab)",
      "grade": "C",
      "fit": "caveat",
      "why": "Thinking switches off through the Qwen3.6 template, but the smallest GGUF is 11.7 GB (IQ2_M) and a 4-bit build about 19 GB, most of a 25 GB spare budget.",
      "source": "https://huggingface.co/Accio-Lab/occamy-1.0/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/occamy-1-0"
    },
    {
      "slug": "ornith-1-5-35b-a3b",
      "name": "Ornith-1.5-35B-A3B",
      "company": "Ornith",
      "grade": "C",
      "fit": "caveat",
      "why": "3B active and enable_thinking false in the template, but the card never mentions the switch and the smallest build is 21.7 GB.",
      "source": "https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B/raw/main/chat_template.jinja",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/ornith-1-5-35b-a3b"
    },
    {
      "slug": "ornith-1-5-9b",
      "name": "Ornith-1.5-9B",
      "company": "Ornith",
      "grade": "C",
      "fit": "caveat",
      "why": "5.8 GB and enable_thinking false in the template, but the card does not document the switch and the model is a long thinker.",
      "source": "https://huggingface.co/ornith-ai/Ornith-1.5-9B/raw/main/chat_template.jinja",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/ornith-1-5-9b"
    },
    {
      "slug": "qwen3-30b-a3b-instruct-2507",
      "name": "Qwen3-30B-A3B-Instruct-2507",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "C",
      "fit": "caveat",
      "why": "No thinking block at all, so routing calls answer at once, but the 4-bit build is 19 GB, most of a 25 GB spare budget.",
      "source": "https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-30b-a3b-instruct-2507"
    },
    {
      "slug": "qwen3-5-9b",
      "name": "Qwen3.5-9B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "C",
      "fit": "caveat",
      "why": "Thinking off per request and Ollama qwen3.5:9b at 6.6 GB, but acc-26 measured it over eight times slower than Granite on routing, past the 3 second bar on a 16 GB GPU.",
      "source": "https://huggingface.co/Qwen/Qwen3.5-9B",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-5-9b"
    },
    {
      "slug": "qwen3-6-35b-a3b",
      "name": "Qwen3.6-35B-A3B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "B",
      "fit": "caveat",
      "why": "acc-36 maps a non-thinking variant to title generation beside the main model, but at 22 GB UD-Q4_K_XL it takes most of a 25 GB spare budget.",
      "source": "https://dzx.fr/blog/local-ai-coding/",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-6-35b-a3b"
    },
    {
      "slug": "qwen3-vl-30b-a3b-instruct",
      "name": "Qwen3-VL-30B-A3B-Instruct",
      "company": "Alibaba Cloud (per the pilot's RDAP check of qwen.ai; this model's files name no copyright holder)",
      "grade": "C",
      "fit": "caveat",
      "why": "No thinking and 3B active, but 18.6 GB at Q4 takes most of a 25 GB spare budget.",
      "source": "https://huggingface.co/api/models/Qwen/Qwen3-VL-30B-A3B-Instruct-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-vl-30b-a3b-instruct"
    },
    {
      "slug": "smollm3-3b",
      "name": "SmolLM3-3B",
      "company": "Hugging Face, Inc.",
      "grade": "C",
      "fit": "caveat",
      "why": "enable_thinking false per request and 1.9 GB Q4 GGUFs from ggml-org and Unsloth, but Hugging Face ships no GGUF and Ollama has no library entry.",
      "source": "https://huggingface.co/HuggingFaceTB/SmolLM3-3B",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/smollm3-3b"
    },
    {
      "slug": "apertus-v1-5-8b",
      "name": "Apertus v1.5 8B",
      "company": "ETH Zurich and EPFL (the Swiss National AI Institute, named in the Acceptable Use Policy)",
      "grade": "C",
      "fit": "no",
      "why": "No publisher GGUF, no Ollama entry and no stock runtime loads the architecture; a cousin cannot be loaded beside a main model.",
      "source": "https://huggingface.co/api/models/swiss-ai/Apertus-v1.5-8B?blobs=true",
      "confidence": "low",
      "url": "https://www.anchorterminal.com/tools/apertus-v1-5-8b"
    },
    {
      "slug": "carbon-x",
      "name": "Carbon-X",
      "company": "Hugging Face (the HuggingFaceBio organisation, named Hugging Face Biology Research on the Hub)",
      "grade": "C",
      "fit": "no",
      "why": "Routing or summary prompts are outside its DNA training; the card's agent section makes it a tool an agent calls, not a cousin that routes.",
      "source": "https://huggingface.co/HuggingFaceBio/Carbon-X",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/carbon-x"
    },
    {
      "slug": "cosmos-reason2-2b",
      "name": "Cosmos-Reason2-2B",
      "company": "NVIDIA Corporation",
      "grade": "C",
      "fit": "no",
      "why": "NVIDIA ships no GGUF and no thinking switch beyond a system-prompt instruction; a routing cousin is not what it is for.",
      "source": "https://raw.githubusercontent.com/nvidia-cosmos/cosmos-reason2/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/cosmos-reason2-2b"
    },
    {
      "slug": "deepseek-v3-2",
      "name": "DeepSeek-V3.2",
      "company": "DeepSeek (the MIT licence names DeepSeek, and DeepSeek's platform terms name Hangzhou DeepSeek Artificial Intelligence Co., Ltd.)",
      "grade": "C",
      "fit": "no",
      "why": "No small build exists; the smallest file is 161 GB.",
      "source": "https://huggingface.co/api/models/unsloth/DeepSeek-V3.2-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/deepseek-v3-2"
    },
    {
      "slug": "deepseek-v4-1-flash",
      "name": "DeepSeek V4.1 Flash",
      "company": "DeepSeek (the MIT licence names DeepSeek, and DeepSeek's platform terms name Hangzhou DeepSeek Artificial Intelligence Co., Ltd.)",
      "grade": "C",
      "fit": "no",
      "why": "No small build; 233 GB minimum.",
      "source": "https://huggingface.co/api/models/unsloth/DeepSeek-V4.1-Flash-1017-GGUF?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/deepseek-v4-1-flash"
    },
    {
      "slug": "deepseek-v4-flash-0731",
      "name": "DeepSeek-V4-Flash-0731",
      "company": "DeepSeek (the MIT licence names DeepSeek, and DeepSeek's platform terms name Hangzhou DeepSeek Artificial Intelligence Co., Ltd.)",
      "grade": "C",
      "fit": "no",
      "why": "Smallest build about 103 GB; it is the main model, not a cousin.",
      "source": "https://unsloth.ai/docs/models/deepseek-v4",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/deepseek-v4-flash-0731"
    },
    {
      "slug": "diffusiongemma-26b-a4b-it",
      "name": "DiffusionGemma 26B A4B IT",
      "company": "Google LLC (RDAP registrant of gemma.ai; the card names Google DeepMind as author)",
      "grade": "D",
      "fit": "no",
      "why": "No Google quantised build, no Ollama entry, and the only local GGUF needs a llama.cpp branch; output arrives one 256-token canvas at a time.",
      "source": "https://huggingface.co/unsloth/diffusiongemma-26B-A4B-it-GGUF",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/diffusiongemma-26b-a4b-it"
    },
    {
      "slug": "glm-4-5-air",
      "name": "GLM-4.5-Air",
      "company": "JINGSHENG HENGXING TECHNOLOGY PTE.LTD (operator of Z.ai per its terms of use, and registrant of z.ai, per the first batch's check on 21 October 2026)",
      "grade": "C",
      "fit": "no",
      "why": "The smallest build anywhere is 38 GB (UD-TQ1_0); it is a main model, not a cousin.",
      "source": "https://huggingface.co/api/models/unsloth/GLM-4.5-Air-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/glm-4-5-air"
    },
    {
      "slug": "glm-5-3",
      "name": "GLM-5.3",
      "company": "JINGSHENG HENGXING TECHNOLOGY PTE.LTD (operator of Z.ai per its terms of use, and registrant of z.ai). The licence names Z.AI",
      "grade": "B",
      "fit": "no",
      "why": "No small build.",
      "source": "https://huggingface.co/api/models/unsloth/GLM-5.3-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/glm-5-3"
    },
    {
      "slug": "glm-5-3-flash",
      "name": "GLM-5.3-Flash",
      "company": "Z.AI Co., Ltd (named in the licence's copyright line)",
      "grade": "B",
      "fit": "no",
      "why": "Smallest build 93 GB.",
      "source": "https://huggingface.co/api/models/unsloth/GLM-5.3-Flash-GGUF?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/glm-5-3-flash"
    },
    {
      "slug": "gpt-oss-120b",
      "name": "gpt-oss-120b",
      "company": "OpenAI (named as publisher on the card; the openai-api listing's terms name OpenAI OpCo, LLC)",
      "grade": "BB",
      "fit": "no",
      "why": "63 GB; it is the main model.",
      "source": "https://huggingface.co/api/models/ggml-org/gpt-oss-120b-GGUF?blobs=true",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gpt-oss-120b"
    },
    {
      "slug": "hy3",
      "name": "Hy3",
      "company": "Tencent (named in the LICENSE file's copyright line)",
      "grade": "B",
      "fit": "no",
      "why": "Smallest build 89 GB.",
      "source": "https://huggingface.co/api/models/AngelSlim/Hy3-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/hy3"
    },
    {
      "slug": "inkling",
      "name": "Inkling",
      "company": "Thinking Machines Lab, Inc. (named on its model card page and in its training-content summary)",
      "grade": "B",
      "fit": "no",
      "why": "No small build.",
      "source": "https://unsloth.ai/docs/models/inkling",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/inkling"
    },
    {
      "slug": "inkling-small",
      "name": "Inkling-Small",
      "company": "Thinking Machines Lab, Inc. (named on its model card page and in its training-content summary)",
      "grade": "B",
      "fit": "no",
      "why": "Smallest build 75 GB.",
      "source": "https://huggingface.co/api/models/unsloth/Inkling-Small-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/inkling-small"
    },
    {
      "slug": "internvl3-5-gpt-oss-20b-a4b-preview",
      "name": "InternVL3.5-GPT-OSS-20B-A4B-Preview",
      "company": "Shanghai AI Laboratory (OpenGVLab), not named in the repository",
      "grade": "D",
      "fit": "no",
      "why": "No publisher GGUF, no Ollama entry, no tool template and no thinking switch.",
      "source": "https://huggingface.co/api/models/OpenGVLab/InternVL3_5-GPT-OSS-20B-A4B-Preview-HF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/internvl3-5-gpt-oss-20b-a4b-preview"
    },
    {
      "slug": "kimi-k3",
      "name": "Kimi K3",
      "company": "Moonshot AI (the licence names Moonshot AI; the Kimi Platform terms name Moonshot AI PTE. LTD.)",
      "grade": "B",
      "fit": "no",
      "why": "No small build.",
      "source": "https://huggingface.co/api/models/unsloth/Kimi-K3-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/kimi-k3"
    },
    {
      "slug": "kolibri-1",
      "name": "Kolibri-1",
      "company": "Aleph Alpha GmbH",
      "grade": "C",
      "fit": "no",
      "why": "Smallest third-party build 17.6 GB at IQ1_S and 49.8 GB at Q4; 78B total is a main model.",
      "source": "https://huggingface.co/api/models/bartowski/Kolibri-1-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/kolibri-1"
    },
    {
      "slug": "kronos",
      "name": "Kronos",
      "company": "No company. Yu Shi publishes it from the GitHub account shiyu-coder and the Hugging Face user account NeoQuasar (named ShiYu), with co-authors at Tsinghua University per the paper",
      "grade": "C",
      "fit": "no",
      "why": "Not a language model; cannot answer routing or summary prompts.",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/kronos"
    },
    {
      "slug": "laguna-m-1",
      "name": "Laguna M.1",
      "company": "Poolside, Inc. (named in Poolside's Terms of Use; the LICENSE.md copyright line reads Poolside)",
      "grade": "C",
      "fit": "no",
      "why": "225B total parameters, 139 GB at the smallest; cannot sit beside a main model in spare memory.",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/laguna-m-1"
    },
    {
      "slug": "laguna-s-2-1",
      "name": "Laguna S 2.1",
      "company": "Poolside, Inc. (named in Poolside's Terms of Use)",
      "grade": "C",
      "fit": "no",
      "why": "96 GB at Q4_K_M; cannot be the second model in spare memory.",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/laguna-s-2-1"
    },
    {
      "slug": "lbs-27b",
      "name": "LBS-27B",
      "company": "inclusionAI, Ant Group (named in the site footer of inclusion-ai.org); the work is credited to authors from Wuhan University of Technology, CASIA, Zhejiang University and Ant Group",
      "grade": "C",
      "fit": "no",
      "why": "BF16 only, 54.7 GB; not a small cousin and not a chat or routing model.",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/lbs-27b"
    },
    {
      "slug": "minimax-m2-5",
      "name": "MiniMax-M2.5",
      "company": "MiniMax AI (named in the licences; MiniMax's platform terms name Nanonoble Pte. Ltd., Singapore)",
      "grade": "C",
      "fit": "no",
      "why": "Smallest build is 56 GB; cannot share memory with a main model.",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/minimax-m2-5"
    },
    {
      "slug": "minimax-m2-7",
      "name": "MiniMax-M2.7",
      "company": "MiniMax (named in the licence's copyright line; MiniMax's platform terms name Nanonoble Pte. Ltd., Singapore)",
      "grade": "C",
      "fit": "no",
      "why": "Smallest build is 61 GB (UD-IQ1_M) and thinking cannot be turned off; it cannot sit beside a main model.",
      "source": "https://huggingface.co/api/models/unsloth/MiniMax-M2.7-GGUF/tree/main?recursive=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/minimax-m2-7"
    },
    {
      "slug": "minimax-m3-1-flash",
      "name": "MiniMax-M3.1-Flash",
      "company": "MiniMax (named in the licence's copyright line; MiniMax's platform terms name Nanonoble Pte. Ltd., Singapore)",
      "grade": "C",
      "fit": "no",
      "why": "A 428B model with a 129 GB smallest build; it cannot be the small cousin.",
      "source": "https://huggingface.co/api/models/unsloth/MiniMax-M3.1-Flash-GGUF/tree/main?recursive=true",
      "confidence": "low",
      "url": "https://www.anchorterminal.com/tools/minimax-m3-1-flash"
    },
    {
      "slug": "mistral-large-4",
      "name": "Mistral Large 4",
      "company": "Mistral AI (RCS Paris 952 418 325)",
      "grade": "C",
      "fit": "no",
      "why": "A 1T-parameter model; no build is small enough to sit beside a main model.",
      "source": "https://huggingface.co/api/models/unsloth/Mistral-Large-4-1T-A52B-GGUF/tree/main?recursive=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/mistral-large-4"
    },
    {
      "slug": "nemotron-3-super-120b-a12b",
      "name": "Nemotron 3 Super 120B A12B",
      "company": "NVIDIA Corporation",
      "grade": "B",
      "fit": "no",
      "why": "Needs about 64 GB to 72 GB at 4 bits; it cannot be the small cousin.",
      "source": "https://unsloth.ai/docs/models/nemotron-3/nemotron-3-super",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-super-120b-a12b"
    },
    {
      "slug": "nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "company": "NVIDIA Corporation",
      "grade": "B",
      "fit": "no",
      "why": "A 550B model; nothing small enough to sit beside a main model.",
      "source": "https://huggingface.co/api/models/unsloth/NVIDIA-Nemotron-3-Ultra-550B-A55B-GGUF/tree/main?recursive=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nemotron-3-ultra-550b-a55b"
    },
    {
      "slug": "olmo-3-7b-think",
      "name": "Olmo 3 7B Think",
      "company": "The Allen Institute for Artificial Intelligence (Ai2), a 501(c)(3) nonprofit, per its Responsible Use Guidelines",
      "grade": "D",
      "fit": "no",
      "why": "Every reply opens a think block with no switch, so a routing or title call cannot skip reasoning.",
      "source": "https://huggingface.co/allenai/Olmo-3-7B-Think/raw/main/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/olmo-3-7b-think"
    },
    {
      "slug": "ornith-1-5-397b",
      "name": "Ornith-1.5-397B",
      "company": "Ornith",
      "grade": "C",
      "fit": "no",
      "why": "397B model.",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/ornith-1-5-397b"
    },
    {
      "slug": "qwen3-5-122b-a10b",
      "name": "Qwen3.5-122B-A10B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "B",
      "fit": "no",
      "why": "At 34 GB even at IQ1_M it is a main model, not a cousin in 2 to 25 GB of spare memory.",
      "source": "https://huggingface.co/api/models/unsloth/Qwen3.5-122B-A10B-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-5-122b-a10b"
    },
    {
      "slug": "qwen3-5-397b-a17b",
      "name": "Qwen3.5-397B-A17B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "B",
      "fit": "no",
      "why": "Over 100 GB at any precision; not a cousin.",
      "source": "https://huggingface.co/api/models/unsloth/Qwen3.5-397B-A17B-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-5-397b-a17b"
    },
    {
      "slug": "qwen3-6-27b",
      "name": "Qwen3.6-27B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "B",
      "fit": "no",
      "why": "A 17 GB Q4 main model; acc-26 measured Qwen 3.6 at 26.7 s on routing prompts, far past 3 seconds.",
      "source": "https://bitworking.org/news/2026/05/surprising-things-i-learned-putting-together-a-home-brain/",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-6-27b"
    },
    {
      "slug": "qwen3-8-27b",
      "name": "Qwen3.8-27B",
      "company": "Alibaba Cloud (named in the licence's copyright line)",
      "grade": "B",
      "fit": "no",
      "why": "A 17 GB Q4 main model with default xhigh reasoning; acc-39 pairs it with a smaller fast model rather than using it as the cousin.",
      "source": "https://huggingface.co/api/models/unsloth/Qwen3.8-27B-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-8-27b"
    },
    {
      "slug": "qwen3-8-flash-next",
      "name": "Qwen3.8-Flash-Next",
      "company": "Alibaba Cloud (Singapore) Private Limited, the registrant of qwen.ai. The licence names Qwen as copyright holder",
      "grade": "C",
      "fit": "no",
      "why": "Over 66 GB at any precision; not a cousin.",
      "source": "https://huggingface.co/api/models/unsloth/Qwen3.8-Flash-Next-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-8-flash-next"
    },
    {
      "slug": "qwen3-coder-next",
      "name": "Qwen3-Coder-Next",
      "company": "Alibaba Cloud (the Qwen team's employer; no LICENSE file names a copyright holder for this model)",
      "grade": "C",
      "fit": "no",
      "why": "48 GB at Q4_K_M; a main model, not a cousin.",
      "source": "https://huggingface.co/api/models/Qwen/Qwen3-Coder-Next-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-coder-next"
    },
    {
      "slug": "qwen3-omni-30b-a3b-instruct",
      "name": "Qwen3-Omni-30B-A3B-Instruct",
      "company": "Alibaba Cloud (carried from the 20 October 2026 pilot; this model's files name no copyright holder)",
      "grade": "D",
      "fit": "no",
      "why": "18.6 GB at Q4 even in ggml-org's GGUF; not a cousin.",
      "source": "https://huggingface.co/api/models/ggml-org/Qwen3-Omni-30B-A3B-Instruct-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-omni-30b-a3b-instruct"
    },
    {
      "slug": "qwen3-vl-32b-instruct",
      "name": "Qwen3-VL-32B-Instruct",
      "company": "Alibaba Cloud (per the first batch's RDAP check of qwen.ai; this model's files name no copyright holder)",
      "grade": "C",
      "fit": "no",
      "why": "19.8 GB at Q4 with 256 KB of KV per token; a main model.",
      "source": "https://huggingface.co/api/models/Qwen/Qwen3-VL-32B-Instruct-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwen3-vl-32b-instruct"
    },
    {
      "slug": "qwopus3-8-27b-flash-v2",
      "name": "Qwopus3.8-27B-Flash-V2",
      "company": "Jackrong, an individual Hugging Face account",
      "grade": "C",
      "fit": "no",
      "why": "A 17 GB Q4 model the card recommends with thinking on at xhigh effort; not a cousin.",
      "source": "https://huggingface.co/Jackrong/Qwopus3.8-27B-Flash-V2",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/qwopus3-8-27b-flash-v2"
    },
    {
      "slug": "recursive-flint",
      "name": "Recursive Flint",
      "company": "Recursive Superintelligence, Inc. (the Hugging Face organisation's full name)",
      "grade": "C",
      "fit": "no",
      "why": "Reasoning always on and 23 GB at INT4; a routing call spends thinking tokens first.",
      "source": "https://huggingface.co/RecursiveSI/flint/raw/main/sglang/README.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/recursive-flint"
    },
    {
      "slug": "ripple-65b-preview",
      "name": "Ripple-65B Preview",
      "company": "Ricursive Intelligence, Inc.",
      "grade": "D",
      "fit": "no",
      "why": "57 GB NVFP4 and always-on thinking; not a cousin.",
      "source": "https://huggingface.co/ricursiveAI/Ripple-65B-Preview",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/ripple-65b-preview"
    },
    {
      "slug": "swift-1-5-qwen3-8-27b",
      "name": "Swift 1.5 Qwen3.8-27B",
      "company": "UkisAI",
      "grade": "C",
      "fit": "no",
      "why": "A 17 GB Q4 model built to think; not a cousin.",
      "source": "https://huggingface.co/api/models/ukisai/Swift-1.5-Qwen3.8-27B-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/swift-1-5-qwen3-8-27b"
    },
    {
      "slug": "swift-1-5-qwen3-8-flash-next",
      "name": "Swift 1.5 Qwen3.8-Flash-Next",
      "company": "UkisAI",
      "grade": "C",
      "fit": "no",
      "why": "Over 66 GB at any precision; not a cousin.",
      "source": "https://huggingface.co/api/models/ukisai/Swift-1.5-Qwen3.8-Flash-Next-GSQ-RCO-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/swift-1-5-qwen3-8-flash-next"
    },
    {
      "slug": "thinkingcap-qwen3-8-27b",
      "name": "ThinkingCap (Qwen3.8 27B)",
      "company": "BottleCap AI (named in the licence's copyright line)",
      "grade": "C",
      "fit": "no",
      "why": "A 17 GB Q4 model tuned for shorter but still-on reasoning; not a cousin.",
      "source": "https://huggingface.co/api/models/bottlecapai/ThinkingCap-Qwen3.8-27B-GGUF?blobs=true",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/thinkingcap-qwen3-8-27b"
    },
    {
      "slug": "unlimited-ocr-2",
      "name": "Baidu Unlimited-OCR-2",
      "company": "Baidu, Inc.",
      "grade": "C",
      "fit": "no",
      "why": "Not a chat model; it has no chat template and answers only page-parsing prompts.",
      "source": "https://huggingface.co/baidu/Unlimited-OCR-2/raw/main/tokenizer_config.json",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/unlimited-ocr-2"
    }
  ],
  "marks": "yes: the docs show it does the job as the success test says. caveat: it does the job with a limit an agent should know. no: it doesn't do this job. unchecked: we couldn't read it.",
  "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": "2026-11-16",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "0.8",
    "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"
  },
  "useCase": {
    "accounts": 7,
    "after": [
      "coding-agents.plan-review"
    ],
    "area": "models",
    "asks": [
      "a cheap model for title generation and routing",
      "something small for the harness's utility calls",
      "a planner and a coder that both fit in memory",
      "a second-opinion reviewer model"
    ],
    "before": [
      "open-weights.local-coding-agent"
    ],
    "blockers": [
      "thinking cannot be turned off, so a routing call takes 20 seconds",
      "no small quantised build, so the cousin takes the main model's memory"
    ],
    "capabilities": [
      "model.open-weights",
      "model.thinking-off",
      "model.gguf-published"
    ],
    "category": "open-weights",
    "categoryName": "Open-weight models",
    "checked": "2026-11-12",
    "companies": {
      "fit": 8,
      "fitWithLimit": 18,
      "tagged": 90
    },
    "competition": "crowded",
    "fits": [
      "gemma-4-12b-it",
      "gemma-4-e2b-it",
      "gemma-4-e4b-it",
      "granite-4-2-3b",
      "ling-3-0-tiny",
      "nemotron-3-5-bolt-e4b-4gb",
      "nemotron-3-nano-4b",
      "qwen3-4b-instruct-2507",
      "qwen3-5-0-8b",
      "qwen3-5-2b",
      "qwen3-5-4b",
      "qwen3-coder-30b-a3b-instruct",
      "qwen3-vl-2b-instruct",
      "qwen3-vl-4b-instruct",
      "qwen3-vl-8b-instruct",
      "spark-x2-6-4b"
    ],
    "fitsWithLimit": [
      "agents-a1-4b",
      "apodex-1-1-mini",
      "bigbang-v1",
      "darwin-36b-opus",
      "gemma-4-26b-a4b-it",
      "gemma-4-31b-it",
      "glm-4-7-flash",
      "gpt-oss-20b",
      "laguna-xs-2-1",
      "lfm2-5-2-6b",
      "mimo-v2-6-distill-qwen-9b",
      "minicpm-o-4-5",
      "minicpm5-2b",
      "muse-glimmer-30b",
      "nanbeige-4-3-33b",
      "nemotron-3-5-lightning-30b-a3b",
      "nemotron-3-nano-30b-a3b",
      "nemotron-3-nano-omni-30b-a3b-reasoning",
      "occamy-1-0",
      "ornith-1-5-35b-a3b",
      "ornith-1-5-9b",
      "qwen3-30b-a3b-instruct-2507",
      "qwen3-5-9b",
      "qwen3-6-35b-a3b",
      "qwen3-vl-30b-a3b-instruct",
      "smollm3-3b"
    ],
    "id": "open-weights.sub-task-cousin",
    "listings": {
      "caveat": 26,
      "no": 44,
      "notRead": 28,
      "tagged": 161,
      "unchecked": 0,
      "yes": 16
    },
    "name": "A small fast cousin for sub-tasks beside a main model",
    "status": "draft",
    "success": "Two models fit co-resident within the machine's memory, the cousin runs with thinking off, and the main model is not reloaded between calls. Measured: the cousin answers a routing or summary prompt correctly in under 3 seconds.",
    "url": "https://www.anchorterminal.com/api/v1/use-cases/open-weights.sub-task-cousin.json",
    "weight": 8,
    "weightBasis": "6 of 51 accounts pair a small model with a larger one for utility or planner roles"
  }
}
