Qwen3 Coder 480B A35B
Qwen · released Jul 22, 2025 · Qwen/Qwen3-Coder-480B-A35B-Instruct
- Type
- Open weightsApache License 2.0
- Params
- 480B
- Context
- 262K
35B active per word · about 197K words of context
Our take
Written Sep 30, 2026Qwen3 Coder 480B A35B is a downloadable coding model you can reach through a host, with a licence that allows commercial use, changes and redistribution. It scores well on real GitHub issues, but human voters place it mid-table on general and coding prompts, and it is far too large to run on your own machine.
Use it for software engineering work where you reach the model through a host and pay by the token, or for tasks that need a long document or a large code file kept in one request without splitting it up. Its request capacity is large enough that long inputs need not be broken apart first. Skip it if you need to run the model on your own machine, or if you want a model that leads on human preference.
The case for it
- 69.6% on SWE-bench Verified via OpenHands, which measures the share of real GitHub issues resolved end-to-end inside that harness rather than general coding ability.
- 262144 tokens of request capacity, so a long document or a large code file need not be split up before you ask about it.
- The licence allows commercial use, changes and redistribution (Apache License 2.0).
The case against it
- Human preference is middling: 88th of 168 on Arena Coding as of 25 Sep 2026, a board recording which answer people preferred rather than whether it was correct.
- Web-app building is among its weaker measured areas, at 83rd of 95 on Arena Code (WebDev) as of 25 Sep 2026.
- 480.2 billion parameters in total, of which 35 billion are active per token, so hosted use is the practical route for nearly everyone.
How good is it?
EverydayGeneral questions and everyday reasoning
Arena Text (overall)107th of 168 · 1387
CodingWriting and fixing code on its own
Arena Coding88th of 168 · 1457
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent. It is on SWE-bench Verified, in 11th of 42 with 69.6.
WritingDrafting and rewriting prose
Arena Creative Writing95th of 168 · 1362
Arena Creative Writing is the only board that has scored it for this.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done.
Every published score for this model8 scoresEvery figure we hold, from 8 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Comfortable fit
Apple M3 Ultra (80-core GPU) · 512 GB
Room to spare. 70.4 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked 4 hours ago — each listing carries its own date.
- per 1M tokens
- $0.30 in / $1.00 out
- Context served
- 262K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $0.30 / $1.00checked 4 hours ago | 262K | not measured | Unknown | Unknown | Unknown |
| DeepInfraturbo tierfp4Through OpenRouter | $0.30 / $1.00checked 4 hours ago | 262K66K max reply | 54 tok/s | No | No | Unknown |
| Venice AIfp8Through OpenRouter | $0.35 / $1.50checked 4 hours ago | 256K66K max reply | 44 tok/s | No | No | Confirmed |
| Novita AIfp8Direct and through OpenRouter | $0.38 / $1.55checked 4 hours ago | 262K66K max reply through OpenRouter | 42 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| Google Vertex AIus-south1Through OpenRouter | $0.22 / $1.80checked 4 hours ago | 262K66K max reply | 68 tok/s | No | No | Confirmed |
| Alibaba CloudThrough OpenRouter | $0.97 / $4.88checked 4 hours ago | 262K66K max reply | 18 tok/s | No | Yesunknown period | Unknown |
DeepInfra
Novita AI
Google Vertex AIAcross the 6 listings we hold: 5 say they do not train on prompts (1 of them only through OpenRouter), 0 say they do and 1 does not say. 3 appear in the zero-retention registry we check (1 of them only through OpenRouter); the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| DeepInfraturbo · fp4Through OpenRouter | ✓ | ✓ | ✓ |
| Venice AIfp8Through OpenRouter | ✓ | ✓ | ✓ |
| Novita AIfp8Direct and through OpenRouter | ✓ | ✓ | ✓ |
| Google Vertex AIus-south1Through OpenRouter | ✓ | ✓ | ✓ |
| Alibaba CloudThrough OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 6 of 6 listings say yes. JSON output: 6 of 6 listings say yes. Strict schema: 6 of 6 listings say yes.
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 6 listings does not say whether it trains on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
Licence and identifiers
What the licence allowsApache License 2.0, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- Qwen/Qwen3-Coder-480B-A35B-Instruct
- Architecture
- Mixture of experts
- Takes in, gives back
- Text in, text out
- Catalogue slug
- qwen-qwen3-coder-480b-a35b