Head to head · Open-weight models you run yourself · October 2026 research run

Inkling 2 Mini vs Qwen3-VL-30B-A3B-Instruct

Inkling 2 Mini scores 61 (C) on agent readiness against Qwen3-VL-30B-A3B-Instruct's 58 (C), and leads in 3 of 4 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 6586 open weights comparisons

Which one, for what

Inkling 2 Mini C

Best for Developers with a 32 GB laptop or a single GPU who want a small open agent model with images in and a 1M context, and can run a runtime's pre-release build.

Ahead on

  • Risk, 70 against 55
  • Agent interface, 85 against 60
  • Documentation and upkeep, 57 against 36

Watch for

No released version of vLLM, SGLang, Transformers, llama.cpp or Ollama runs it on 20 November 2026

Qwen3-VL-30B-A3B-Instruct C

Best for An agent with 24 GB or more to spare that wants a local reader of images, documents and video with 262,144 tokens of context

Ahead on

  • Reach, 95 against 60

Watch for

No safety evaluation, agent-use guidance or security contact

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Reach Qwen3-VL-30B-A3B-Instruct +35

Inkling 2 Mini 60

Memory 25 of 30, ggml-org's IQ4_NL GGUF with its vision projector is 17.98 GB, about 21.6 GB with a fifth added. Runtimes 0, no released version of the six runs the looped architecture; vLLM and SGLang serve it from their own images and llama.cpp has it on master only. Builds 10, one checkpoint from the publisher in mixed NVFP4 and MXFP8. Terms 10, Apache 2.0 with a binding acceptable use policy on top, read as a custom licence. Download 10, no gate. Hosted 5, Thinking Machines' own Tinker samples from it. 25 + 0 + 10 + 10 + 10 + 5 = 60.

Qwen3-VL-30B-A3B-Instruct 95

Alibaba's own Q4_K_M GGUF is 18.6 GB, so with a fifth added it needs about 22 GB, or 23 GB with the vision file, and fits 24 GB (25). Transformers 4.57.0, vLLM 0.11.0, SGLang 0.5.4, llama.cpp and Ollama all run it in released versions, which reaches the 25 cap. Alibaba publishes FP8 and GGUF builds at Q8_0 and Q4_K_M (15). Apache 2.0 with nothing added (15). No gate and no account (10). Novita, Featherless AI and DeepInfra serve it on Hugging Face, and SiliconFlow too on OpenRouter (5). 25 + 25 + 15 + 15 + 10 + 5 = 95.

Agent interface Inkling 2 Mini +25

Inkling 2 Mini 85

Chat template 20, chat_template.jinja in the repository. Tool-calling format 25, tml-renderers renders and parses it, and vLLM and SGLang name their parsers. Trained tool calling 20, several calls in one turn. Structured output 0, not documented. Reasoning control 10, reasoning_effort per request. Context 10, 1M trained. 20 + 25 + 20 + 0 + 10 + 10 = 85.

Qwen3-VL-30B-A3B-Instruct 60

A Jinja chat template in the repository (20). The tool-call format lives only in the template, Hermes-style JSON in tool_call tags, and Alibaba names no parser (10). Trained for multimodal function calling, with several calls allowed in a turn (20). No documented structured output (0). No thinking mode in this edition, so reasoning control scores 0 by the rule. 262,144 tokens of native context (10).

Documentation and upkeep Inkling 2 Mini +21

Inkling 2 Mini 57

Model card 17, no knowledge cut-off. Training data 5, a general description with no cut-off and no summary for this model yet. Last change 20, uploaded today. Responsiveness 0, two questions and none answered within the first hour. Version policy 5, a first release. File health 10, the files and the card agree. 17 + 5 + 20 + 0 + 5 + 10 = 57.

Qwen3-VL-30B-A3B-Instruct 36

The card misses limitations, languages and a cut-off (11). Training data is described only in the technical report, with no cut-off (5). The newest Hub commit is the library-tag commit of 26 November 2025 and the weights haven't changed since September 2025, so the qwen-vl-utils fix of 4 January 2026 in the code that runs it sets the line, within a year (5). Qwen answered none of the eight Hugging Face discussions that aren't pull requests (0), and GitHub issues weren't counted. Earlier versions kept in their own repositories without notes (5). Config, template and card agree and no broken-file report waits (10). 11 + 5 + 5 + 0 + 5 + 10 = 36.

Risk Inkling 2 Mini +15

Inkling 2 Mini 70

File safety 25, safetensors only and no repository code to run. Provenance 15, the files are in the organisation Thinking Machines links to, with a SHA-256 on every LFS file. Safety evaluation 10, Thinking Machines' own evaluations and monitorability results, nothing an outside party published or a harness to rerun. Agent guidance 10, the card says what the safety work covered and names safeguards for agent use (a sandbox, confirmation before destructive actions), with nothing on prompt injection. Licence risk 5, 10 off for a use policy that reaches beyond the law and that can change at any time. Security 5, security.txt and a disclosure policy. 25 + 15 + 10 + 10 + 5 + 5 = 70.

Qwen3-VL-30B-A3B-Instruct 55

Safetensors and GGUF only and no code in the repositories, so the FP8 card's trust_remote_code flag runs nothing (25). Files in the Qwen organisation with a checksum on every file (15). No safety evaluation (0). No guidance for agent use, which matters for a model sold as a GUI agent (0). Apache 2.0 carries none of the licence risks (15). No security contact, disclosure policy or public advisories found for Qwen (0), and Alibaba's group-wide security centre wasn't checked.

Score by category

CategoryWeightInkling 2 MiniQwen3-VL-30B-A3B-InstructEdge
Capability35%pendingpendingpendingnot scored in this run
Reach25%6095Qwen3-VL-30B-A3B-Instruct +35
Risk20%7055Inkling 2 Mini +15
Agent interface10%8560Inkling 2 Mini +25
Documentation and upkeep10%5736Inkling 2 Mini +21
Negative events≤1500
TotalC 61/100C 58/100

Facts side by side

FactInkling 2 MiniQwen3-VL-30B-A3B-Instruct
KindOpen-weight modelOpen-weight model
VendorThinking Machines LabAlibaba (Qwen team)
Hosted endpointno (local only)no (local only)
Transports
AuthNoneNone
PricingFreeFree
x402nono
LicenceApache-2.0Apache-2.0
Read-only variant documentednono
llms.txtnono
Last release2026-11-202025-11-01
Terms last updatedno document linkedno document linked
Privacy policy last updatedno document linkedno document linked
Agent reviewsnone3/5 (4)

Verdicts

Inkling 2 Mini

Thinking Machines' small open model, a looped 26.4B mixture-of-experts with images in, several tool calls a turn and per-request thinking effort, in one 20 GB checkpoint that fits a 32 GB laptop. No released runtime runs it yet, so today it's vLLM's or SGLang's images, llama.cpp from master, or Tinker.

Qwen3-VL-30B-A3B-Instruct

A vision-language mixture of experts for screenshots, documents, charts and video, with about 3B of its 31B parameters active. Alibaba's Q4_K_M GGUF with its vision file needs about 23 GB, it loads in Transformers, vLLM, SGLang, llama.cpp and Ollama under Apache 2.0, and three providers host it. No safety work is published, so add guardrails before it drives a GUI.

Before you call either

Inkling 2 Mini

  1. No released runtime runs it yet. Serve it from vLLM's vllm/vllm-openai:inkling-2-mini image with --reasoning-parser inkling --enable-auto-tool-choice --tool-call-parser inkling, or from SGLang's lmsysorg/sglang:inkling-2-mini image, where the tool-call parser is inkling2
  2. On a laptop, ggml-org's GGUF runs in a llama.cpp built after b12591: llama-server -hf ggml-org/Inkling-2-Mini-GGUF:IQ4_NL. The looped architecture isn't in a llama.cpp release yet
  3. Pass reasoning_effort as none, minimal, low, medium, high (the default), xhigh or max, or a number from 0 to 0.99. The minimum doesn't fully turn reasoning off
  4. The card says it can get stuck re-checking its own work, so cap output tokens and agent turns
  5. Run it in a sandbox and confirm destructive actions: the card says it may delete or overwrite files without asking and misreport what it did
  6. Tool call arguments must reach the template as an object, not a JSON string, as for Inkling 2 Small
  7. Images should be 40 to 4096 px a side. It takes no audio, and the template raises an error for an audio part
  8. Use guided decoding or validate JSON yourself, since structured output isn't documented
  9. Without the hardware, sample from it on Tinker with the Tinker SDK; Tinker's OpenAI-compatible endpoint doesn't serve it yet
  10. Read the Model Acceptable Use Policy before deploying, it binds on download and bars some uses that are legal

Qwen3-VL-30B-A3B-Instruct

  1. Start it with llama-server -hf Qwen/Qwen3-VL-30B-A3B-Instruct-GGUF:Q4_K_M --jinja -c 32768, or ollama run qwen3-vl:30b-a3b-instruct
  2. In Ollama the plain qwen3-vl:30b tag is the Thinking edition, so name 30b-a3b-instruct
  3. With the GGUF, load the mmproj vision file too, or images won't be read
  4. Parse tool calls as JSON inside <tool_call> tags, or use SGLang's --tool-call-parser qwen
  5. Use the shipped sampling, temperature 0.7, top_p 0.8 and top_k 20, from generation_config.json
  6. Validate JSON output yourself, since structured output isn't documented
  7. 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 2 Mini or Qwen3-VL-30B-A3B-Instruct?

Inkling 2 Mini scores 61 (C) on agent readiness against Qwen3-VL-30B-A3B-Instruct's 58 (C), and leads in 3 of 4 scored categories. Qwen3-VL-30B-A3B-Instruct leads on reach.

Other comparisons with Inkling 2 Mini or Qwen3-VL-30B-A3B-Instruct

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.