qwen
Qwen: Qwen3.6 27B
qwen/qwen3.6-27b
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Qwen3.6-27B is a 27-billion-parameter open-weight model from Alibaba's Qwen team, released in April 2026 under an Apache-2.0 license according to its model card. It pairs a hybrid linear-attention and gated-attention architecture with a vision encoder, accepting text and image inputs; the OpenRouter catalog additionally lists video input. It offers a 262,144-token context, extensible to about one million tokens per the card, optional reasoning with preserved thinking context, tool calling, and structured outputs. The vendor positions it for agentic coding, and its self-reported benchmarks emphasize coding-agent tasks.[1][2]
Strengths
- Long context: 262,144 tokens natively, with the model card claiming extension up to roughly one million tokens.[1][2]
- Stated focus on agentic coding: the card reports gains in frontend workflows and repository-level reasoning, and the catalog includes an Artificial Analysis coding index.[1][2]
- Optional reasoning enabled by default in the OpenRouter listing, with a card-described option to retain reasoning context across messages, plus tool and structured-output support.[1][2]
Limitations
- Text-only output; no image, audio, or video generation is reported in either source.[1][2]
- Evidence is mostly vendor-supplied: benchmark figures come from Qwen's own card with only limited third-party indices in the catalog, so validate on your own workload before adoption.[1][2]
- Video input appears only in the OpenRouter catalog while the model card tags the pipeline as image-text-to-text; confirm modality support before relying on video. Knowledge cutoff is not stated.[1][2]
Where to use it
- Agentic coding and repository-level assistance, matching the vendor's stated focus; worth piloting rather than assuming fit.[2]
- Question answering over long documents or image-grounded inputs that benefit from the large context and vision encoder.[1][2]
- Self-hosted deployment via Transformers, vLLM, SGLang, or KTransformers for teams wanting open-weight control.[2]
Sources
- [1] OpenRouter catalog · Qwen: Qwen3.6 27BSources checked: Nov 13, 2026
- [2] Model documentation · Qwen: Qwen3.6 27BSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026. It features hybrid multimodal capabilities — accepting text, image, and video inputs...
Specifications
- Model ID
- qwen/qwen3.6-27b
- Canonical ID
- qwen/qwen3.6-27b-20260422
- Creator
- qwen
- Added to OpenRouter
- 27.04.2026
- Context length
- 262,144
- Modality
- text+image+video->text
- Input modalities
- text, image, video
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.320 / out $3.2
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | reasoning: optional · parameters: include_reasoning, reasoning |
| Coding | BB — next 30% of the cohort | — |
| Agents | BB — next 30% of the cohort | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: optional · parameters: include_reasoning, reasoning |
| Document analysis | No measurement | — |
| Vision | No measurement | input: image |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | input: video |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Coding index | 53.7 | BB — next 30% of the cohort | 13.11.2026 |
| Artificial Analysis | Agentic index | 18.5 | BB — next 30% of the cohort | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Chutesfp8 | 262,144 | $0.250 | $2 | — | — | 100% |
| SiliconFlowfp8 | 262,144 | $0.300 | $3.2 | — | — | 100% |
| DeepInfrafp8 | 262,144 | $0.320 | $3.2 | — | — | 100% |
| Venicefp8 | 256,000 | $0.325 | $3.25 | — | — | 100% |
| Alibaba | 262,144 | $0.450 | $2.7 | — | — | 90% |
| Reka | 262,144 | $0.600 | $6 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.