Qwen3 Coder 30B A3B Instruct
Qwen · released Jul 31, 2025 · Qwen/Qwen3-Coder-30B-A3B-Instruct
- Type
- Open weightsApache License 2.0
- Params
- 30.5B
- Context
- 262K
3B active per word · about 197K words of context
Our take
Written Sep 3, 2026Qwen3 Coder is a downloadable coding specialist with 30.5 billion parameters and only 3 billion active per token, built for software-engineering tasks. It resolves real GitHub issues end-to-end in testing and carries a permissive Apache licence.
Pick this for local or self-hosted coding assistance where a small active footprint and measured code-fixing ability matter. Use it for budget API coding with strong throughput, or for long-context code review across large repositories. Skip it if you need general chat quality scores, multimodal input, or guaranteed fast inference from any provider.
The case for it
- Resolves real GitHub issues end-to-end: 60.4% on SWE-bench Verified.
- Only 3 billion parameters active per token from 30.5 billion total — roughly a 10:1 sparsity ratio.
- Apache 2.0 licence allows commercial use, fine-tuning and redistribution.
- Wide price range with a very cheap entry point among tracked offers.
The case against it
- Throughput varies more than tenfold by provider, from fast to very slow.
- No measured general chat or reasoning quality; only coding benchmarks are listed.
- The vendor's own hosted rate is several times higher than the cheapest tracked offer.
How good is it?
EverydayGeneral questions and everyday reasoning
Not yet scored on Arena Text (overall).
CodingWriting and fixing code on its own
Not yet scored on Arena Coding.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent. It is on SWE-bench Verified, in 23rd of 42 with 60.4.
WritingDrafting and rewriting prose
Not yet scored on Arena Creative Writing.
Every published score for this model1 scoreEvery figure we hold, from 1 board, 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
Borderline fit on an estimated size. It leaves 1.6 GB spare on a size we calculated rather than measured, and a 10% error either way would change the answer.
Comfortable fit
GeForce RTX 5090 · 32 GB
Room to spare. 9.6 GB spare means a 10% error in the size would not change the answer.
Apple M1 Pro (16-core GPU) · 32 GB
Room to spare. 2.8 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.
Cheapest of the 2 listings we can compare like for like — at 262K of context, out of 5 in the table below. One cheaper row there is outside that comparison: a different context length.
- per 1M tokens
- $0.070 in / $0.28 out
- Context served
- 262K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| Novita AIfp8Direct and through OpenRouter | $0.070 / $0.27checked 4 hours ago | 160K33K max reply through OpenRouter | 45 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| OpenRouterOpenRouter's own listing | $0.070 / $0.28checked 4 hours ago | 262K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8Through OpenRouter | $0.070 / $0.28checked 4 hours ago | 262K236K max reply | 23 tok/s | No | No | Confirmed |
| Amazon BedrockThrough OpenRouter | $0.15 / $0.60checked 4 hours ago | 0 | not measured | No | No | Confirmed |
| Alibaba CloudThrough OpenRouter | $0.29 / $1.46checked 4 hours ago | 262K66K max reply | 39 tok/s | No | Yesunknown period | Unknown |
Novita AI
Amazon BedrockAcross the 5 listings we hold: 4 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 |
|---|---|---|---|
| Novita AIfp8Direct and through OpenRouter | ✓ | ✓ | ✗ |
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| SiliconFlowfp8Through OpenRouter | ✓ | ✓ | ✓ |
| Amazon BedrockThrough OpenRouter | ✓ | ✗ | ✗ |
| Alibaba CloudThrough OpenRouter | ✓ | ✓ | ✗ |
Tool calling: 5 of 5 listings say yes. JSON output: 4 of 5 listings say yes, 1 says no. Strict schema: 2 of 5 listings say yes, 3 say no.
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 5 listings does not say whether it trains on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no cached-input rate for any of its listings.
- 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-30B-A3B-Instruct
- Architecture
- Mixture of experts
- Takes in, gives back
- Text in, text out
- Catalogue slug
- qwen-qwen3-coder-30b-a3b-instruct