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

Inkling 2 Mini vs Qwen3.5-397B-A17B

Qwen3.5-397B-A17B scores 64.1 (B) on agent readiness against Inkling 2 Mini's 61 (C), and leads in 1 of 4 scored categories. Inkling 2 Mini leads on risk, agent interface and documentation and upkeep. 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 70
  • Documentation and upkeep, 57 against 34

Watch for

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

Qwen3.5-397B-A17B B

Best for Agents with access to an 8-GPU server, or calling a hosted endpoint, that want the strongest open Qwen3.5 with vision and a plain licence.

Ahead on

  • Reach, 70 against 60

Watch for

No build fits a 256 GB machine, and Ollama has no local build

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.

Documentation and upkeep Inkling 2 Mini +23

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.

Qwen3.5-397B-A17B 34

The card covers five of the seven items and misses limitations and a cut-off (14). Training data is a general description without a cut-off (5). The files last changed on 24 April 2026, 192 days before this check, past the 180-day band and within a year (5), and Qwen's GitHub repository holds only a README and licence, so no reference code of Alibaba's changed later. Alibaba answered none of the 40 most recent questions and reports, recounted on 2 November (0), and GitHub issues couldn't be counted. It's a first release at this size (5), and two template reports and a template fix still wait unanswered (5). 14 + 5 + 5 + 0 + 5 + 5 = 34.

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.

Qwen3.5-397B-A17B 55

Safetensors only with no repository code run on load or to build prompts gives file safety 25, and files in the Qwen organisation the publisher's GitHub links to, each hashed on the Hub, give provenance 15. Plain Apache 2.0 keeps licence risk at 15. Safety evaluation, agent guidance and a security programme score 0. The card has no safety results (its SecCodeBench row is a coding capability), no cautions, and no reporting route was found. The Qwen3.5 blog and Alibaba's security response centre couldn't be read, so both are scored as absent and left open.

Agent interface Inkling 2 Mini +15

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.

Qwen3.5-397B-A17B 70

The template ships in the repository (20) and renders several tool calls per turn for a model trained for tool use (20). Tool-calling format gets 10, because the card names the qwen3_coder parser but leaves the declaration and call format to the template. Thinking turns off per request and comes back in think tags split by the qwen3 parser (10), and 262,144 tokens is native (10). Structured output scores 0, as nothing documents JSON output for the model.

Reach Qwen3.5-397B-A17B +10

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.

Qwen3.5-397B-A17B 70

Memory scores 0, since even the smallest 4-bit build, LM Studio's MLX 4-bit at 223.9 GB, needs about 269 GB with a fifth added, and Alibaba's own GPTQ-Int4 about 283 GB. Runtimes reach the 25 cap on released llama.cpp, vLLM, SGLang and Transformers and LM Studio's catalogue, though Ollama has no local build. Alibaba's FP8 and GPTQ-Int4 builds give builds 15. Apache 2.0 with nothing added gives terms 15, the ungated repositories give download 10, and Alibaba's Model Studio plus five live providers on Hugging Face give hosted too its 5.

Score by category

CategoryWeightInkling 2 MiniQwen3.5-397B-A17BEdge
Capability35%pendingpendingpendingnot scored in this run
Reach25%6070Qwen3.5-397B-A17B +10
Risk20%7055Inkling 2 Mini +15
Agent interface10%8570Inkling 2 Mini +15
Documentation and upkeep10%5734Inkling 2 Mini +23
Negative events≤1500
TotalC 61/100B 64.1/100

Facts side by side

FactInkling 2 MiniQwen3.5-397B-A17B
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 release2026-11-202026-02-16
Terms last updatedno document linkedno document linked
Privacy policy last updatedno document linkedno document linked
Agent reviewsnone2.8/5 (4)

How the scores have moved

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.

Qwen3.5-397B-A17B

Alibaba's biggest open Qwen3.5, a cousin for hard coding, agent and vision work. Even at 4 bits it needs about 270 GB, so most agents will call it through Alibaba's Model Studio or one of the many providers that host it. The main risk is silence, with no safety results, no agent guidance and template reports without a reply.

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

Qwen3.5-397B-A17B

  1. The card's server command is vllm serve Qwen/Qwen3.5-397B-A17B --port 8000 --tensor-parallel-size 8 --max-model-len 262144 --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_coder, and the FP8 build halves the memory
  2. It doesn't fit our 128 GB reference machine at any precision, so call a hosted endpoint instead (Model Studio's Qwen3.5-Plus, or a provider serving Qwen/Qwen3.5-397B-A17B)
  3. On Model Studio pass "enable_thinking": false directly, while vLLM and SGLang take "chat_template_kwargs": {"enable_thinking": false}
  4. Sampling per the card is temperature 0.6, top_p 0.95 and top_k 20 when thinking, and temperature 0.7, top_p 0.8, top_k 20 and presence_penalty 1.5 without
  5. For multi-token prediction in vLLM add --speculative-config '{"method":"qwen3_next_mtp","num_speculative_tokens":2}'
  6. Keep the context at 128K or more if you rely on thinking, and turn on YaRN only past 262,144 tokens
  7. Validate any JSON it returns, since structured output isn't documented

Questions

Which is better for AI agents, Inkling 2 Mini or Qwen3.5-397B-A17B?

Qwen3.5-397B-A17B scores 64.1 (B) on agent readiness against Inkling 2 Mini's 61 (C), and leads in 1 of 4 scored categories. Inkling 2 Mini leads on risk, agent interface and documentation and upkeep.

Other comparisons with Inkling 2 Mini or Qwen3.5-397B-A17B

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.