Qwen instruction model for multilingual chat, reasoning, and tool use
Specification
most-agreed values
Context
262K
Max output
236K
Released
2026-02-23
Knowledge cutoff
2025-04
Retires
—
Open weights
yes
Input
text, image, video
Output
text
Price
US dollars per million tokens · most-agreed
Input
$0.10
Output
$0.15
Cache read
$0.025
Cache write
$0.30
Quality
3 benchmarks · 1 source
Composite
48th
percentile of 334 scored models
Rank
174 / 334
±0.10 sd
Evidence
3 × 1
one source, or fewer than 3 benchmarks
Effort range
—
one configuration only
reasoning56th2/4 bench
math41st1/7 bench
Every score
one row per source and configuration — nothing averaged away
| Benchmark | Score | Configuration | Source | Run |
|---|---|---|---|---|
| GPQA diamond | 79.0 ±2.4 | default | Epoch AI | 2026-08-27 |
| OTIS Mock AIME 2024-2025 | 42.8 ±6.5 | default | Epoch AI | 2026-08-27 |
| Chess Puzzles | 2.0 ±1.4 | default | Epoch AI | 2026-08-27 |
The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.
Available from 23 hosts
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| qvac | free | free | — | — | 33K | 8K | — |
| crof | $0.04 | $0.15 | $0.008 | — | 262K | 262K | — |
| nano-gpt | $0.05 | $0.15 | $0.025 | — | 256K | 66K | — |
| empiriolabs | $0.09 | $0.13 | $0.045 | — | 262K | 33K | — |
| merge-gateway | $0.09 | $0.13 | — | — | 262K | 33K | — |
| deepinfra | $0.10 | $0.15 | — | — | 262K | 66K | — |
| kilo | $0.10 | $0.15 | — | — | 262K | 236K | — |
| llmgateway | $0.10 | $0.15 | — | — | 262K | 66K | — |
| llmgateway-providers | $0.10 | $0.15 | — | — | 262K | 33K | — |
| mixlayer | $0.10 | $0.40 | — | — | 262K | 262K | — |
| openrouter | $0.10 | $0.15 | — | — | 262K | 236K | — |
| siliconflow | $0.10 | $0.15 | — | — | 262K | 262K | — |
| venice | $0.10 | $0.15 | — | — | 256K | 33K | — |
| cortecs | $0.111 | $0.167 | — | — | 262K | 262K | — |
| ovhcloud | $0.12 | $0.18 | — | — | 262K | 262K | — |
| regolo-ai | $0.15 | $0.60 | — | — | 262K | 8K | — |
| tensorx | $0.15 | $0.20 | $0.0375 | $0.1875 | 262K | 66K | — |
| routing-run | $0.16 | $0.48 | — | — | 262K | 32K | — |
| huggingface | $0.17 | $0.25 | — | — | 262K | 66K | — |
| together_ai | $0.17 | $0.25 | — | — | 262K | — | — |
| togetherai | $0.17 | $0.25 | — | — | 262K | 66K | — |
| siliconflow-cn | $0.22 | $1.74 | — | — | 262K | 66K | — |
| pioneer | $0.30 | $0.30 | $0.30 | $0.30 | 33K | 33K | — |
Price history
append-only observations · a listing writes a row only when its price moves
| Date | Host | Field | $/M |
|---|---|---|---|
| 2026-09-02 | openrouter | Input | $0.10 |
| 2026-09-06 | openrouter | Input | $0.10 |
| 2026-09-02 | openrouter | Output | $0.15 |
| 2026-09-06 | openrouter | Output | $0.15 |
More from Alibaba
most-hosted first