AlibabaQwen3 Coder 480B
Best configuration ranks #321 of 610 on the BenchLeader Index at 48.5 ±3.9. Last measured 8 Sept 2026. Released 22 Jul 2025.
- Blended price
- $1.51/M
- $0.861 in · $3.44 out
- Output speed
- 53 tok/s
- First answer
- 2.97 s
- first token 1.05 s
- Context
- 262k
- Overall index49
- Reasoning52
- Coding46
- Agents & tools43
- Knowledge54
- Instruction following43
- Human preference57
- Long context49
- Composite44
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source | Trend |
|---|---|---|---|
| LMArena Hard Prompts | 1414#139 | LMArena | |
| GPQA Diamond (AA)not in index | 61.8%#346 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.5%#414 | Artificial Analysis | |
| Kagi LLM Benchmark | 49.5%#85 | Kagi LLM Benchmark |
Coding
| Benchmark | default | Source | Trend |
|---|---|---|---|
| WeirdML | 41.2%#95 | WeirdML | |
| GSO-Bench | 4.9%#21 | GSO-Bench | |
| LMArena Coding | 1458#119 | LMArena | |
| LMArena WebDev | 1274#101 | LMArena | |
| SWE-Bench Pro | 38.7%#13 | Scale AI SEAL | |
| SWE-bench Verified (bash only) | 55.4%#28 | SWE-bench | |
| SWE-bench Verified (any scaffold)not in index | 55.4%#36 | SWE-bench |
Agents & tools
| Benchmark | default | Source | Trend |
|---|---|---|---|
| Terminal-Bench | 27.2%#47 | Terminal-Bench | |
| Terminal-Bench Hard | 18.9%#161 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 43.6%#208 | Artificial Analysis |
Knowledge
| Benchmark | default | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -22.2#188 | Artificial Analysis |
Instruction following
| Benchmark | default | Source | Trend |
|---|---|---|---|
| IFBench | 40.5%#250 | Artificial Analysis |
Human preference
| Benchmark | default | Source | Trend |
|---|---|---|---|
| LMArena Text | 1387#146 | LMArena |
Long context
| Benchmark | default | Source | Trend |
|---|---|---|---|
| AA-LCR | 45.7%#266 | Artificial Analysis |
Composite
| Benchmark | default | Source | Trend |
|---|---|---|---|
| AA Intelligence Index | 11.9#273 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Google Vertex | 49 tok/s | 1.05 s | $0.220 | $1.80 | 262k | – |
| Venice | 39 tok/s | 0.93 s | $0.350 | $1.50 | 256k | fp8 |
| NovitaAI | 34 tok/s | 0.98 s | $0.380 | $1.55 | 262k | fp8 |
| DeepInfra (Turbo) | 33 tok/s | 1.05 s | $0.300 | $1.00 | 262k | fp4 |
| Alibaba OpenSource | 30 tok/s | 1.37 s | $0.975 | $4.88 | 262k | – |
Price history
Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0014 | – | 8.6 s |
| Summarise a 30-page report | 12,000 / 600 | $0.012 | – | 14.3 s |
| Code edit | 6,000 / 1,500 | $0.010 | – | 31.3 s |
| Agentic coding session | 60,000 / 4,000 | $0.065 | – | 1.3 min |
| Structured extraction | 2,000 / 200 | $0.0024 | – | 6.7 s |
See also
Data as of 9 Sept 2026. Compare with another model.