Qwen3 Coder 480B A35B
AlibabaAlibaba's large Qwen3 model tuned specifically for coding, built as a mixture-of-experts architecture.
Main uses: Code generation and agentic coding tasks at open-model pricing, text-only.
What sets it apart: A 480B-parameter model, named explicitly for its "A35B" active-parameter MoE routing - large by design, not a naming artifact.
In
$0.30
Out
$1.00
Blended
$0.48
Where to buy this model
2 access routes| Route | In | Out | Blended | vs |
|---|---|---|---|---|
| $0.30 | $1.00 | $0.48 | - | |
| ● Venice fp8 | $0.35 | $1.50 | $0.64 | +33% |
● official API · ● cloud platform · ● inference host · USD per 1M tokens. Same model, different providers - price and quantization vary.
- Ctx
- 262K
- max output
- 66K
- cache read /1M
- $0.10
- Family
- qwen
- Modalities
- text
- Features
- json · tools
Live pricing, updated 2026-11-03.
Alibaba models at a similar price
| Model | In | Out | Ctx |
|---|---|---|---|
| Qwen3.6 Flash | $0.188 | $1.13 | 1M |
| Qwen3 VL 30B A3B Thinking | $0.29 | $1 | 262K |
| Qwen3 Coder 480B A35B · this model | $0.3 | $1 | 262K |
| Qwen3.7 Plus | $0.32 | $1.28 | 1M |
| Qwen3.8 27B | $0.15 | $1.88 | 1.024M |
| Qwen3.5 Plus 2026-02-15 | $0.26 | $1.56 | 1M |