Qwen3 Coder 30B A3B Instruct pricing & benchmarks
Released Jul 31, 2025 · served by 5 providers
What Qwen3 Coder 30B A3B Instruct is
Qwen3-Coder-30B-A3B-Instruct is a 30.5B parameter Mixture-of-Experts (MoE) model with 128 experts (8 active per forward pass), designed for advanced code generation, repository-scale understanding, and agentic tool use. Built on the...
- Model ID:
qwen/qwen3-coder-30b-a3b-instruct - Modality: text->text
- Knowledge cutoff: 2025-06-30
- Tool calling: supported · Structured output: supported
- Weights:
Qwen/Qwen3-Coder-30B-A3B-Instructon Hugging Face
Providers and prices for Qwen3 Coder 30B A3B Instruct
The same model costs different amounts depending on who serves it. Prices are per 1M tokens, cheapest first. Uptime and throughput are OpenRouter's rolling measurements, not vendor claims.
| Provider | Input | Output | Context | Quant | Uptime 24h | Throughput |
|---|---|---|---|---|---|---|
| Novitacheapest | $0.070 | $0.270 | 160K | fp8 | 94.5% | — |
| SiliconFlow | $0.070 | $0.280 | 262K | fp8 | 97.4% | — |
| Amazon Bedrock | $0.150 | $0.600 | — | — | — | — |
| Alibaba | $0.292 | $1.46 | 262K | — | 99.5% | — |
| DigitalOcean | $0.450 | $1.70 | 262K | — | 99.5% | — |
Qwen3 Coder 30B A3B Instruct benchmark results
Design Arena head-to-head results, as reported through the OpenRouter model API.
| Category | Arena | Rank | Elo | Win rate |
|---|---|---|---|---|
| uicomponent | models | #81 | 1083 | 54.1% |
| dataviz | models | #82 | 1113 | 54.7% |
| website | models | #85 | 1112 | 57.1% |
FAQ
How much does the Qwen3 Coder 30B A3B Instruct API cost?
$0.070 per 1M input tokens and $0.270 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $0.15.
Is Qwen3 Coder 30B A3B Instruct free?
No. It is a paid model starting at $0.070 per 1M input tokens, though some providers offer trial credits.
Which provider is cheapest for Qwen3 Coder 30B A3B Instruct?
Novita at $0.070 input / $0.270 output per 1M tokens (fp8 quantization).
Can I self-host Qwen3 Coder 30B A3B Instruct?
Yes — weights are published as Qwen/Qwen3-Coder-30B-A3B-Instruct on Hugging Face, so you can run it on your own hardware.