qwen
Qwen: Qwen3 Coder 480B A35B
qwen/qwen3-coder
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Qwen's Qwen3-Coder-480B-A35B-Instruct, listed as qwen/qwen3-coder, is a Mixture-of-Experts code model with 480B total and 35B activated parameters, released under Apache-2.0. The vendor positions it for agentic coding, with a purpose-built function-call format, tool use, and a 256K-token context extendable to 1M via Yarn. OpenRouter's catalog confirms text-to-text operation, tool and structured-output parameters, and publishes Design Arena ratings in coding categories; independent benchmark coverage remains limited.[1][2]
Strengths
- Vendor-reported agentic coding focus: a purpose-built function-call format, tool calling via OpenAI-compatible APIs, and support on agent platforms such as Qwen Code and CLINE.[1][2]
- Long context for repository-scale work: 262,144 tokens natively, extendable to 1M with Yarn per the model card.[1][2]
- Sparse MoE design (160 experts, 8 active; 35B of 480B parameters activated), Apache-2.0 licensing, and catalog support for structured outputs and common sampling controls.[1][2]
Limitations
- Text-only input and output; no vision or other modalities are listed in either source.[1][2]
- Non-thinking mode only: the model card states it does not produce think blocks, and the catalog reports no reasoning capability.[1][2]
- Performance claims are largely vendor-reported; with evidence limited to catalog data and Design Arena ratings, teams should validate quality, latency, and memory needs on their own workloads.[1][2]
Where to use it
- Agentic coding assistants and IDE or CLI integrations that rely on function calling and tool use.[1][2]
- Repository-scale code understanding, refactoring, and review where the long context window may help.[2]
- Web and UI code generation or prototyping, areas covered by the catalog's Design Arena categories.[1]
Sources
- [1] OpenRouter catalog · Qwen: Qwen3 Coder 480B A35BSources checked: Nov 13, 2026
- [2] Model documentation · Qwen: Qwen3 Coder 480B A35BSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over...
Specifications
- Model ID
- qwen/qwen3-coder
- Canonical ID
- qwen/qwen3-coder-480b-a35b-07-25
- Creator
- qwen
- Added to OpenRouter
- 23.07.2025
- Context length
- 262,144
- Modality
- text->text
- Input modalities
- text
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.300 / out $1
- Knowledge cutoff
- 30.06.2025
- 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 | — |
| Coding | No measurement | — |
| Agents | No measurement | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | — |
| Document analysis | No measurement | — |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Design Arena | models · codecategorieswin rate 61.2% · rank 97 | 1,133Elo | CC — below the B band | 13.11.2026 |
| Design Arena | models · datavizwin rate 54.9% · rank 108 | 1,084Elo | CC — below the B band | 13.11.2026 |
| Design Arena | models · gamedevwin rate 58.7% · rank 100 | 1,089Elo | CC — below the B band | 13.11.2026 |
| Design Arena | models · uicomponentwin rate 61.4% · rank 98 | 1,105Elo | CC — below the B band | 13.11.2026 |
| Design Arena | models · websitewin rate 61.7% · rank 95 | 1,151Elo | CC — below the B band | 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) |
|---|---|---|---|---|---|---|
| DeepInfrafp4 | 262,144 | $0.300 | $1 | — | — | 65% |
| Venicefp8 | 256,000 | $0.350 | $1.5 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.