qwen/qwen3-coder
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...
| Provider | In $/M | Out $/M | Context | Uptime |
|---|---|---|---|---|
| $0.220 | $1.80 | 262K | 100% | |
| DeepInfrafp4 | $0.300 | $1.00 | 262K | 98.9% |
| Venicefp8 | $0.350 | $1.50 | 256K | — |
| Novitafp8 | $0.380 | $1.55 | 262K | 100% |
| AtlasCloudfp8 | $0.780 | $3.80 | 262K | — |
| Alibaba | $0.975 | $4.88 | 262K | — |
| WandBbf16 | $1.00 | $1.50 | 262K | — |
Qwen3 Coder 480B A35B costs $0.220 per million input tokens and $1.80 per million output tokens via OpenRouter, making it 105th cheapest of 298 paid models.
Qwen3 Coder 480B A35B scores 24.8 on the Artificial Analysis Intelligence Index, ranking 97th of 180 benchmarked models, with a GPQA Diamond score of 62%.
Qwen3 Coder 480B A35B generates around 45 tokens per second with 654ms time-to-first-token (p50), the 193rd fastest tracked model.
Qwen3 Coder 480B A35B supports a 1.0M-token context window and can output up to 66K tokens. It accepts text input.