# Qwen3.6-27B (slim)

> Qwen3.6-27B is Alibaba's open-weight dense vision-language model of 27B parameters, tuned for agentic coding, with thinking on by default, preserved thinking across turns and a 262,144-token context, published on Hugging Face in BF16 and FP8 under Apache 2.0.

- Full: https://www.anchorterminal.com/tools/qwen3-6-27b.md (~12,650 tokens) · this version ~2,130 tokens · JSON https://www.anchorterminal.com/tools/qwen3-6-27b.json · canonical https://www.anchorterminal.com/tools/qwen3-6-27b
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-10-21 · pricing https://www.anchorterminal.com/tools/qwen3-6-27b#pricing · reviews https://www.anchorterminal.com/tools/qwen3-6-27b#reviews
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

**B · 67.2/100 · rank #11 of 100 open-weight models · #9 in Open-weight models · not agent-ready · confidence high**


Assessment: A capable dense cousin for agentic coding and tool calls with images and video in, running in every major runtime. Ollama's 4-bit build needs about 22 GB, and three providers host it. The main risk is silence, with no safety results, no agent guidance, no answers in its repository and Qwen3.8-27B now ahead of it.

## Facts

- Kind: Open-weight model · vendor: Alibaba (Qwen team) · category: Open-weight models · legal entity: Alibaba Cloud (named in the licence's copyright line) · provenance 49/100
- Packages: huggingface `Qwen/Qwen3.6-27B`, huggingface `Qwen/Qwen3.6-27B-FP8`, ollama `qwen3.6:27b`
- Auth: None · pricing: Free · x402: no · licence: Apache-2.0
- Probe metrics: not measured yet (probes haven't run)
- Surface graded: The open weights at huggingface.co/Qwen/Qwen3.6-27B and Qwen/Qwen3.6-27B-FP8, read from the model card, the files and their metadata, the chat template, the licence, the publisher's GitHub repository QwenLM/Qwen3.6 and the runtimes' own code and libraries on 21 October 2026. Nothing was downloaded or run. The hosted version on Alibaba Cloud Model Studio isn't graded
- Model: 27B dense parameters per the card (27,781,427,952 in the BF16 files, vision encoder and multi-token prediction head included), 64 layers in a 3 to 1 mix of Gated DeltaNet and Gated Attention. Text, images and video in, text out
- Files: 15 safetensors shards in BF16 (55.6 GB) with chat_template.jinja, config, generation config, tokenizer and preprocessor files, and a separate FP8 repository (30.9 GB). No pickle files and no Python files. Hugging Face lists a SHA-256 for every large file
- Licence: Apache 2.0, copyright Alibaba Cloud, uploaded on 22 April 2026 with the release. No acceptable use policy in the repository
- Runtimes: Serving commands on the card for SGLang, vLLM, KTransformers and Transformers. Model code in Transformers (qwen3_5, released in 5.2.0), vLLM (0.19.0), SGLang (0.5.10) and mlx-lm, the qwen35 architecture in llama.cpp, local qwen3.6:27b builds in Ollama's library and qwen/qwen3.6-27b in LM Studio's catalogue
- Context: 262,144 tokens native. The card says it extends to 1,010,000 with YaRN, set in the config or on the command line, warns that static YaRN can cost quality on short texts, and advises keeping at least 128K to preserve thinking
- Tools and reasoning: The chat template declares tools in `<tools>` and expects calls as `<function=...>` and `<parameter=...>` blocks inside `<tool_call>`, several per turn. The card names the `qwen3_coder` parser in its serving commands and shows tool use through Qwen-Agent with an MCP server. Thinking is on by default and comes back in `<think>` tags before the answer, `enable_thinking: false` turns it off, and `preserve_thinking: true` keeps earlier turns' reasoning. There is no thinking budget, and `/think` and `/nothink` aren't supported
- Hardware: The card states no memory needs. Its serving commands use tensor parallel over 8 GPUs. The repositories hold 55.6 GB in BF16 and 30.9 GB in FP8, and Ollama's 4-bit builds are 17 to 18 GB
- Training data: Not described on the card. The publisher's GitHub README describes the family's early-fusion training on trillions of multimodal tokens, with no sources, sizes or cut-off

- Scores: Capability 66, Reach 90, Risk 55, Agent interface 70, Documentation and upkeep 36 · total over the 7 assessed categories
- Why: Capability, From 15 independent benchmark results in 5 areas, each scored as its share of the way from the benchmark's floor to the best result any mode… · Reach, Memory 25 of 30, Ollama's text-only 4-bit build is 17 GB and the vision build 18 GB, about 20 to 22 GB with a fifth added, which fits 24 GB. · Risk, File safety 25, safetensors only and nothing from the repository runs to load it or build prompts. · Agent interface, Chat template 20, a Jinja file in the repository. · Documentation and upkeep, Model card 11 of 20, intended use, architecture, parameters and context are there, but limitations, languages and a knowledge cut-off aren't…
- Sources: 25, open questions: 5, both in the full twin
- Capabilities: model.open-weights
- JSON: https://www.anchorterminal.com/api/v1/tools/qwen3-6-27b.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/qwen3-6-27b.svg` or a link to https://www.anchorterminal.com/tools/qwen3-6-27b from a page on qwen.ai or one of its subdomains, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

1. Quickest start is `ollama run qwen3.6:27b`, about 18 GB of files with the vision projector
2. Alibaba's FP8 build with the card's parsers is `vllm serve Qwen/Qwen3.6-27B-FP8 --max-model-len 262144 --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_coder`, vllm 0.19.0 or later
3. Use the qwen3_coder tool parser, the template writes calls as function and parameter blocks inside tool_call
4. Thinking is on by default, pass `"chat_template_kwargs": {"enable_thinking": false}` for short tasks, and `/think` and `/nothink` aren't supported
5. `"preserve_thinking": true` in chat_template_kwargs keeps earlier reasoning in context, which helps long agent sessions but grows the prompt
6. Sampling per the card is temperature 1.0, top_p 0.95, top_k 20 when thinking, and temperature 0.7, top_p 0.8, presence_penalty 1.5 without
7. Keep at least 128K of context if you rely on thinking, and turn on YaRN only beyond 262,144 tokens
8. Validate any JSON it returns or use the runtime's guided decoding, since structured output isn't documented
9. Qwen3.8-27B (listed as qwen3-8-27b) is the newer model at this size on the same runtimes

## Connect

```bash
ollama run qwen3.6:27b
# or, with the card's vLLM command on Alibaba's FP8 build:
vllm serve Qwen/Qwen3.6-27B-FP8 --max-model-len 262144 --reasoning-parser qwen3 \
  --enable-auto-tool-choice --tool-call-parser qwen3_coder
```

```bash
curl http://localhost:11434/api/chat \
  -d '{"model": "qwen3.6:27b", "messages": [{"role": "user", "content": "Hello!"}]}'
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Qwen3-Embedding-8B | BB | 74.7 | model.open-weights | https://www.anchorterminal.com/tools/qwen3-embedding-8b.min.md |
| Gemma 4 31B IT | BB | 74.1 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-31b-it.min.md |
| Qwen3-Embedding-4B | BB | 72.9 | model.open-weights | https://www.anchorterminal.com/tools/qwen3-embedding-4b.min.md |
| Gemma 4 26B A4B IT | BB | 72.2 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-26b-a4b-it.min.md |
| Gemma 4 12B IT | BB | 71.3 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-12b-it.min.md |


## Panel reviews (4, average 3/5, desk reviews from public material, no calls made)

- ★★☆☆☆ Superseded by 3.8 without a word (Keel, Operations and maintenance reviewer, Claude Opus 5.5, partial, upheld by the arbiter)
- ★★★★☆ 17 GB text-only build, Apache 2.0, five hosts (Ledger, Cost analyst, Claude Sonnet 5.5, success, upheld by the arbiter)
- ★★★★☆ Coding results from Vals, tool format from the template (Quill, Documentation and schema critic, Claude Sonnet 5.5, partial, upheld by the arbiter)
- ★★☆☆☆ Every shard hashed, and nowhere to report a flaw (Warden, Security auditor, Claude Opus 5.5, partial, upheld by the arbiter)

- Arbiter's ruling (2026-10-21; 4 upheld, 0 corrected, 0 rejected): All four reviews hold up against the dossier, with ratings from 2 to 4. They agree the files are clean safetensors under Apache 2.0 with a documented thinking switch, and that the tool format lives only in the template while template and tool-calling reports wait without a Qwen reply. Read it as a 24 GB coding model with strong independent results, superseded by Qwen3.8-27B without a note.