Head to head · Open-weight models you run yourself · October 2026 research run
Inkling vs Qwen3-VL-30B-A3B-Instruct
Inkling scores 64.2 (B) on agent readiness against Qwen3-VL-30B-A3B-Instruct's 58 (C), and leads in 4 of 5 scored categories. Qwen3-VL-30B-A3B-Instruct leads on reach. Both do open-weight models you run yourself.
Best open-weight models you download and run yourself for AI agents · All 2870 open weights comparisons
InklingThinking Machines Lab · Open-weight model
B64.2/100#20 of 100 open-weight models
Qwen3-VL-30B-A3B-InstructAlibaba (Qwen team) · Open-weight model
C58/100#68 of 100 open-weight models
Which one, for what
Inkling B
Best for Agents calling a hosted open multimodal model with audio input, or teams with a multi-GPU datacentre node who want to run it themselves.
Ahead on
- Capability, 60 against 39
- Risk, 70 against 55
- Agent interface, 85 against 60
- Documentation and upkeep, 82 against 36
Watch for
The smallest build Thinking Machines publishes needs about 710 GB, beyond any single workstation
Best for An agent with 24 GB or more to spare that wants a faster, stronger local reader of images, documents and video than the 8B model
Ahead on
- Reach, 95 against 50
Watch for
No safety evaluation, agent-use guidance or security contact
Score by category
| Category | Weight this run | Inkling | Qwen3-VL-30B-A3B-Instruct | Edge |
|---|---|---|---|---|
| Capability | 35%35 | 60 | 39 | Inkling +21 |
| Reach | 25%25 | 50 | 95 | Qwen3-VL-30B-A3B-Instruct +45 |
| Risk | 20%20 | 70 | 55 | Inkling +15 |
| Agent interface | 10%10 | 85 | 60 | Inkling +25 |
| Documentation and upkeep | 10%10 | 82 | 36 | Inkling +46 |
| Negative events | ≤15 | 0 | 0 | |
| Total | B 64.2/100 | C 58/100 |
Facts side by side
| Fact | Inkling | Qwen3-VL-30B-A3B-Instruct |
|---|---|---|
| Kind | Open-weight model | Open-weight model |
| Vendor | Thinking Machines Lab | Alibaba (Qwen team) |
| Hosted endpoint | no (local only) | no (local only) |
| Transports | ||
| Auth | None | None |
| Pricing | Free | Free |
| x402 | no | no |
| Licence | Apache-2.0 | Apache-2.0 |
| Read-only variant documented | no | no |
| llms.txt | no | no |
| Last release | 2026-07-23 | 2025-11-01 |
| Terms last updated | no document linked | no document linked |
| Privacy policy last updated | no document linked | no document linked |
| Customer content may train models | ||
| Terms restrict automated access | ||
| Terms restrict benchmarking | ||
| Terms or service can change without notice | ||
| Arbitration or class-action waiver | ||
| Agent reviews | 3/5 (4) | 3/5 (4) |
Verdicts
Inkling
Thinking Machines' large multimodal generalist, stronger than Inkling-Small on factual recall by its own numbers. It needs about 710 GB of datacentre GPU memory even in NVFP4, so most agents will use a host such as Together AI or Tinker. The main risks are a use policy that binds on download and can change, and no outside safety results.
Qwen3-VL-30B-A3B-Instruct
A mid-sized vision-language cousin for screenshots, documents, charts and video that runs fast for its size, with about 3B active parameters. It needs about 23 GB for Alibaba's Q4_K_M GGUF with its vision file, loads in every major runtime under plain Apache 2.0, and three providers host it. With no published safety work, add guardrails before it drives a GUI.
Before you call either
Inkling
- Unless you have a multi-GPU Blackwell or H200 node, call it through Together AI, Fireworks, DeepInfra or Tinker rather than starting it
- On 8 B200 GPUs, SGLang's recipe is
sglang serve --trust-remote-code --model-path thinkingmachines/Inkling-NVFP4 --tp 8 --quantization modelopt_fp4 --enable-multimodal --reasoning-parser inkling --tool-call-parser inkling - Use the inkling tool parser, not Hermes, and the inkling reasoning parser
- Pass
reasoning_effortas none, minimal, low, medium, high (the default) or max, or a number from 0 to 0.99. The repository's template has no xhigh preset and raises an error for it, though Thinking Machines' docs list xhigh - Tool call arguments must reach the template as an object, not a JSON string, or it raises an error
- Audio should be 16 kHz WAV, ideally under 20 minutes, and images 40 to 4096 px a side, per the card
- Read the Model Acceptable Use Policy before deploying, it binds on download and bars some uses that are legal
Qwen3-VL-30B-A3B-Instruct
- Start it with
llama-server -hf Qwen/Qwen3-VL-30B-A3B-Instruct-GGUF:Q4_K_M --jinja -c 32768, orollama run qwen3-vl:30b-a3b-instruct - In Ollama the plain
qwen3-vl:30btag is the Thinking edition, so name30b-a3b-instruct - With the GGUF, load the mmproj vision file too, or images won't be read
- Parse tool calls as JSON inside
<tool_call>tags, or use SGLang's--tool-call-parser qwen - Use the shipped sampling, temperature 0.7, top_p 0.8 and top_k 20, from generation_config.json
- Validate JSON output yourself, since structured output isn't documented
- vLLM needs 0.11.0 or later, Transformers 4.57.0 or later and llama.cpp b6907 or later
Questions
Which is better for AI agents, Inkling or Qwen3-VL-30B-A3B-Instruct?
Inkling scores 64.2 (B) on agent readiness against Qwen3-VL-30B-A3B-Instruct's 58 (C), and leads in 4 of 5 scored categories. Qwen3-VL-30B-A3B-Instruct leads on reach.
Other comparisons with Inkling or Qwen3-VL-30B-A3B-Instruct
- Agents-A1-4B vs Inkling
- Agents-A1-4B vs Qwen3-VL-30B-A3B-Instruct
- Apertus v1.5 8B vs Inkling
- Apertus v1.5 8B vs Qwen3-VL-30B-A3B-Instruct
- Apodex-1.1-mini vs Inkling
- Apodex-1.1-mini vs Qwen3-VL-30B-A3B-Instruct
- BigBang-v1 vs Inkling
- BigBang-v1 vs Qwen3-VL-30B-A3B-Instruct
- Cosmos-Reason2-2B vs Inkling
- Cosmos-Reason2-2B vs Qwen3-VL-30B-A3B-Instruct
- Darwin-36B-Opus vs Inkling
- Darwin-36B-Opus vs Qwen3-VL-30B-A3B-Instruct
- DeepSeek-V3.2 vs Inkling
- DeepSeek-V3.2 vs Qwen3-VL-30B-A3B-Instruct
- DeepSeek V4.1 Flash vs Inkling
- DeepSeek V4.1 Flash vs Qwen3-VL-30B-A3B-Instruct
- DeepSeek-V4-Flash-0731 vs Inkling
- DeepSeek-V4-Flash-0731 vs Qwen3-VL-30B-A3B-Instruct
- DiffusionGemma 26B A4B IT vs Inkling
- DiffusionGemma 26B A4B IT vs Qwen3-VL-30B-A3B-Instruct
- Gemma 4 12B IT vs Inkling
- Gemma 4 12B IT vs Qwen3-VL-30B-A3B-Instruct
- Gemma 4 26B A4B IT vs Inkling
- Gemma 4 26B A4B IT vs Qwen3-VL-30B-A3B-Instruct
- Gemma 4 31B IT vs Inkling
- Gemma 4 31B IT vs Qwen3-VL-30B-A3B-Instruct
- Gemma 4 E2B IT vs Inkling
- Gemma 4 E2B IT vs Qwen3-VL-30B-A3B-Instruct
- Gemma 4 E4B IT vs Inkling
- Gemma 4 E4B IT vs Qwen3-VL-30B-A3B-Instruct
- GLM-4.5-Air vs Inkling
- GLM-4.5-Air vs Qwen3-VL-30B-A3B-Instruct
- GLM-4.7-Flash vs Inkling
- GLM-4.7-Flash vs Qwen3-VL-30B-A3B-Instruct
- GLM-5.3-Flash vs Inkling
- GLM-5.3-Flash vs Qwen3-VL-30B-A3B-Instruct
- GLM-5.3 vs Inkling
- GLM-5.3 vs Qwen3-VL-30B-A3B-Instruct
- gpt-oss-120b vs Inkling
- gpt-oss-120b vs Qwen3-VL-30B-A3B-Instruct
- gpt-oss-20b vs Inkling
- gpt-oss-20b vs Qwen3-VL-30B-A3B-Instruct
- Granite 4.2 3B vs Inkling
- Granite 4.2 3B vs Qwen3-VL-30B-A3B-Instruct
- Hy3 vs Inkling
- Hy3 vs Qwen3-VL-30B-A3B-Instruct
- Inkling-Small vs Qwen3-VL-30B-A3B-Instruct
- Inkling vs Inkling-Small
- Inkling vs InternVL3.5-GPT-OSS-20B-A4B-Preview
- Inkling vs Kimi K3
- Inkling vs Kolibri-1
- Inkling vs Laguna M.1
- Inkling vs Laguna S 2.1
- Inkling vs Laguna XS 2.1
- Inkling vs LFM2.5-2.6B
- Inkling vs Ling-3.0-tiny
- Inkling vs MiMo-V2.6 Distill Qwen 9B
- Inkling vs MiniCPM-o 4.5
- Inkling vs MiniCPM5-2B
- Inkling vs MiniMax-M2.5
- Inkling vs MiniMax-M2.7
- Inkling vs Muse Glimmer 30B
- Inkling vs Nemotron 3.5 Lightning 30B A3B
- Inkling vs Nemotron 3 Nano 30B A3B
- Inkling vs Nemotron 3 Nano 4B
- Inkling vs Nemotron 3 Nano Omni 30B A3B Reasoning
- Inkling vs Nemotron 3 Super 120B A12B
- Inkling vs Nemotron 3 Ultra 550B A55B
- Inkling vs occamy-1.0
- Inkling vs Olmo 3 7B Think
- Inkling vs Ornith-1.5-35B-A3B
- Inkling vs Ornith-1.5-397B
- Inkling vs Ornith-1.5-9B
- Inkling vs Qwen3-30B-A3B-Instruct-2507
- Inkling vs Qwen3-4B-Instruct-2507
- Inkling vs Qwen3.5-0.8B
- Inkling vs Qwen3.5-122B-A10B
- Inkling vs Qwen3.5-2B
- Inkling vs Qwen3.5-397B-A17B
- Inkling vs Qwen3.5-4B
- Inkling vs Qwen3.5-9B
- Inkling vs Qwen3.6-27B
- Inkling vs Qwen3.6-35B-A3B
- Inkling vs Qwen3.8-27B
- Inkling vs Qwen3.8-Flash-Next
- Inkling vs Qwen3-Coder-30B-A3B-Instruct
- Inkling vs Qwen3-Coder-Next
- Inkling vs Qwen3-Omni-30B-A3B-Instruct
- Inkling vs Qwen3-VL-2B-Instruct
- Inkling vs Qwen3-VL-32B-Instruct
- Inkling vs Qwen3-VL-4B-Instruct
- Inkling vs Qwen3-VL-8B-Instruct
- Inkling vs Qwopus3.8-27B-Flash-V2
- Inkling vs SmolLM3-3B
- Inkling vs Spark-X2.6-4B
- Inkling vs Swift 1.5 Qwen3.8-27B
- Inkling vs Swift 1.5 Qwen3.8-Flash-Next
- Inkling vs ThinkingCap (Qwen3.8 27B)
- InternVL3.5-GPT-OSS-20B-A4B-Preview vs Qwen3-VL-30B-A3B-Instruct
- Kimi K3 vs Qwen3-VL-30B-A3B-Instruct
- Kolibri-1 vs Qwen3-VL-30B-A3B-Instruct
- Laguna M.1 vs Qwen3-VL-30B-A3B-Instruct
- Laguna S 2.1 vs Qwen3-VL-30B-A3B-Instruct
- Laguna XS 2.1 vs Qwen3-VL-30B-A3B-Instruct
- LFM2.5-2.6B vs Qwen3-VL-30B-A3B-Instruct
- Ling-3.0-tiny vs Qwen3-VL-30B-A3B-Instruct
- MiMo-V2.6 Distill Qwen 9B vs Qwen3-VL-30B-A3B-Instruct
- MiniCPM-o 4.5 vs Qwen3-VL-30B-A3B-Instruct
- MiniCPM5-2B vs Qwen3-VL-30B-A3B-Instruct
- MiniMax-M2.5 vs Qwen3-VL-30B-A3B-Instruct
- MiniMax-M2.7 vs Qwen3-VL-30B-A3B-Instruct
- Muse Glimmer 30B vs Qwen3-VL-30B-A3B-Instruct
- Nemotron 3.5 Lightning 30B A3B vs Qwen3-VL-30B-A3B-Instruct
- Nemotron 3 Nano 30B A3B vs Qwen3-VL-30B-A3B-Instruct
- Nemotron 3 Nano 4B vs Qwen3-VL-30B-A3B-Instruct
- Nemotron 3 Nano Omni 30B A3B Reasoning vs Qwen3-VL-30B-A3B-Instruct
- Nemotron 3 Super 120B A12B vs Qwen3-VL-30B-A3B-Instruct
- Nemotron 3 Ultra 550B A55B vs Qwen3-VL-30B-A3B-Instruct
- occamy-1.0 vs Qwen3-VL-30B-A3B-Instruct
- Olmo 3 7B Think vs Qwen3-VL-30B-A3B-Instruct
- Ornith-1.5-35B-A3B vs Qwen3-VL-30B-A3B-Instruct
- Ornith-1.5-397B vs Qwen3-VL-30B-A3B-Instruct
- Ornith-1.5-9B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-30B-A3B-Instruct-2507 vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-4B-Instruct-2507 vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.5-0.8B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.5-122B-A10B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.5-2B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.5-397B-A17B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.5-4B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.5-9B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.6-27B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.6-35B-A3B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.8-27B vs Qwen3-VL-30B-A3B-Instruct
- Qwen3.8-Flash-Next vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-Coder-30B-A3B-Instruct vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-Coder-Next vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-Omni-30B-A3B-Instruct vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-VL-2B-Instruct vs Qwen3-VL-30B-A3B-Instruct
- Qwen3-VL-30B-A3B-Instruct vs Qwen3-VL-32B-Instruct
- Qwen3-VL-30B-A3B-Instruct vs Qwen3-VL-4B-Instruct
- Qwen3-VL-30B-A3B-Instruct vs Qwen3-VL-8B-Instruct
- Qwen3-VL-30B-A3B-Instruct vs Qwopus3.8-27B-Flash-V2
- Qwen3-VL-30B-A3B-Instruct vs SmolLM3-3B
- Qwen3-VL-30B-A3B-Instruct vs Spark-X2.6-4B
- Qwen3-VL-30B-A3B-Instruct vs Swift 1.5 Qwen3.8-27B
- Qwen3-VL-30B-A3B-Instruct vs Swift 1.5 Qwen3.8-Flash-Next
- Qwen3-VL-30B-A3B-Instruct vs ThinkingCap (Qwen3.8 27B)
Machine-readable
- This page as Markdown
/compare/inkling-vs-qwen3-vl-30b-a3b-instruct.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/inkling.json·/api/v1/tools/qwen3-vl-30b-a3b-instruct.json - From a terminal
anchor compare inkling qwen3-vl-30b-a3b-instruct(the CLI) - Over MCP
compare_tools {"a": "inkling", "b": "qwen3-vl-30b-a3b-instruct"}at/mcp, no key