AlibabaQwen3 Omni 30B A3B Instruct
Best configuration ranks #482 of 610 on the BenchLeader Index at 41.6 ±5.5 (thinking reasoning effort). Released 24 Sept 2025.
- Blended price
- $0.430/M
- $0.250 in · $0.970 out
- Output speed
- 100 tok/s
- First answer
- 22 s
- first token 1.86 s
- Context
- 66k
- Overall index42
- Agents & tools37
- Knowledge35
- Instruction following45
- Multimodal44
- Long context25
- Composite39
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| thinkingbest | 41.6 | #482 | 100 tok/s | 22 s | $0.0004 | Agents & tools 37 · Composite 39 · Instruction following 45 · Knowledge 35 · Long context 25 · Multimodal 44 |
| default | 39.0 | #537 | 96 tok/s | 1.86 s | $0.0004 | Agents & tools 35 · Composite 37 · Instruction following 35 · Knowledge 31 · Long context 25 · Multimodal 40 |
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 | thinking | default | Source |
|---|---|---|---|
| GPQA Diamond (AA)not in index | 72.6%#267 | 62.0%#345 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 7.5%#293 | 4.6%#406 | Artificial Analysis |
Agents & tools
| Benchmark | thinking | default | Source |
|---|---|---|---|
| Terminal-Bench Hard | 3.8%#283 | 1.5%#321 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 21.4%#309 | 16.4%#339 | Artificial Analysis |
Knowledge
| Benchmark | thinking | default | Source |
|---|---|---|---|
| AA-Omniscience | -61.4#397 | -69.3#433 | Artificial Analysis |
Instruction following
| Benchmark | thinking | default | Source |
|---|---|---|---|
| IFBench | 43.4%#219 | 31.2%#339 | Artificial Analysis |
Multimodal
| Benchmark | thinking | default | Source |
|---|---|---|---|
| MMMU-Pro | 60.2%#177 | 55.5%#192 | Artificial Analysis |
Long context
| Benchmark | thinking | default | Source |
|---|---|---|---|
| AA-LCR | 0.0%#419 | 0.0%#419 | Artificial Analysis |
Composite
| Benchmark | thinking | default | Source |
|---|---|---|---|
| AA Intelligence Index | 7.8#382 | 6.0#459 | Artificial Analysis |
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 | $0.0004 | – | 24.9 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0036 | – | 27.8 s |
| Code edit | 6,000 / 1,500 | $0.0030 | – | 36.8 s |
| Agentic coding session | 60,000 / 4,000 | $0.019 | – | 1.0 min |
| Structured extraction | 2,000 / 200 | $0.0007 | – | 23.9 s |
See also
Data as of 9 Sept 2026. Compare these configurations.