AlibabaQwen3 Coder 480B
Qwen3 Coder 480B is an Alibaba open-weights model, released 22 Jul 2025. Its best configuration ranks #439 of 785 on the BenchLeader Index at 48.0 ±3.6, in the lower half. The ± is the point: 297 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (56) and lowest in agents & tools (42). At $0.635 per million tokens blended it is mid-priced. Output speed of 82 tokens per second puts it slower than most, with a first answer in 1.1 s. Last measured 23 Oct 2026.
- Blended price
- $0.635/M
- $0.380 in · $1.40 out
- Output speed
- 82 tok/s
- measured by Artificial Analysis
- First answer
- 1.13 s
- first token 0.62 s
- Context
- 262k
- Full answer
- 7.22 s
- median, reasoning included
- Cost per run
- –
- Released
- 22 Jul 2025
- Overall index48
- Reasoning46
- Coding46
- Agents & tools42
- Maths55
- Knowledge52
- Instruction following43
- Human preference56
- Long context47
- Composite43
How its index has moved
8 Sept 2026 to 23 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. 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 | not stated | Source | Trend |
|---|---|---|---|
| Arena Hard Prompts | 1414#159 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 61.8%#363 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.5%#493 | Artificial Analysis | |
| Kagi LLM Benchmark | 49.5%#85 | Kagi LLM Benchmark | |
| CritPt | 0.0%#315 | Artificial Analysis |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| WeirdML | 41.2%#97 | WeirdML | |
| GSO-Bench | 4.9%#25 | GSO-Bench | |
| Arena Coding | 1457#139 | Arena (LMArena) | |
| Arena WebDev | 1275#121 | Arena (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 | 69.6%#15 | SWE-bench | |
| SWE-Bench Pro (private)not in index | 38.7%#13 | Scale AI SEAL | |
| ALE-Benchnot in index | 461.4#105 | Epoch AI Benchmarking Hub | |
| AlgoTunenot in index | 1.4#13 | AlgoTune |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench | 27.2%#63 | Terminal-Bench | |
| Terminal-Bench Hard | 18.9%#166 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 43.6%#214 | Artificial Analysis |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Maths | 1376#176 | Arena (LMArena) |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -22.2#250 | Artificial Analysis | |
| AA-Omniscience: accuracynot in index | 15.7%#405 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 55.1%#99 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 40.5%#259 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1383#157 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1388#165 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1362#156 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1397#156 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1408#148 | Arena (LMArena) |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR | 45.7%#335 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index v4.3.2 | 11.9#340 | 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 |
|---|---|---|---|---|---|---|
| DeepInfra (Turbo) | 74 tok/s | 0.41 s | $0.300 | $1.00 | 262k | fp4 |
| Venice | 10 tok/s | 1.13 s | $0.350 | $1.50 | 256k | fp8 |
Price history
The price above, per 1M tokens, every day it changed.
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 | $0.0006 | – | 4.8 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0054 | – | 8.4 s |
| Code edit | 6,000 / 1,500 | $0.0044 | – | 19.4 s |
| Agentic coding session | 60,000 / 4,000 | $0.028 | – | 49.9 s |
| Structured extraction | 2,000 / 200 | $0.0010 | – | 3.6 s |
See also
Questions people ask
- What are Qwen3 Coder 480B's benchmark scores?
- As of 25 Oct 2026, Qwen3 Coder 480B scores 48.0 ±3.6 on the BenchLeader Index, strongest in human preference (56) and weakest in agents & tools (42). That rests on 17 independent benchmark results, and every score on this page links to its publisher.
- How does Qwen3 Coder 480B rank?
- Qwen3 Coder 480B is #439 of 785 ranked configurations on 25 Oct 2026. 297 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Qwen3 Coder 480B cost per 1M tokens?
- As of 25 Oct 2026, Qwen3 Coder 480B costs $0.380 per 1M input tokens and $1.40 per 1M output tokens, $0.635 blended (three input tokens per output token). The list price last changed on 24 Oct 2026.
- When was Qwen3 Coder 480B released?
- Alibaba released Qwen3 Coder 480B on 22 Jul 2025.
Data as of 25 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Qwen3 Coder 480B: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/qwen3-coder-480b.