AlibabaQwen2.5 7B Instruct
Not yet ranked: too few independent results so far. Released 19 Sept 2024.
- Blended price
- $0.090/M
- $0.072 in · $0.144 out
- Output speed
- 51 tok/s
- First answer
- 0.26 s
- Context
- 131k
- Reasoning28
- Maths25
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 |
|---|---|---|
| GPQA Diamond | 35.5%#239 | Epoch AI Benchmarking Hub |
Maths
| Benchmark | default | Source |
|---|---|---|
| OTIS Mock AIME | 2.5%#236 | Epoch AI Benchmarking Hub |
Composite
| Benchmark | default | Source |
|---|---|---|
| Epoch Capabilities Indexnot in index | 118.5#169 | Epoch AI Benchmarking Hub |
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 |
|---|---|---|---|---|---|---|
| Phala | 51 tok/s | 0.26 s | $0.100 | $0.200 | 33k | – |
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 | $0.0001 | – | 6.1 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0010 | – | 12.0 s |
| Code edit | 6,000 / 1,500 | $0.0006 | – | 29.7 s |
| Agentic coding session | 60,000 / 4,000 | $0.0049 | – | 1.3 min |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 4.2 s |
See also
Data as of 9 Sept 2026. Compare with another model.