qwen3-5
Also known as
Qwen3 5 is available through 62 API providers on LMSpeed. Compare API pricing from $0.0005 to $225.00 per million input tokens across providers. Free API access is offered by 1 provider. In speed benchmarks, the fastest provider reaches 165 tok/s.
Compare Qwen3 5 API pricing across 61 providers. Input prices range from $0.0005 to $225.00 per million input. 91VIP offers the lowest rate at $0.0005/M.
| Provider | Model Variant | Input ($/M) | Output ($/M) | Speed (t/s) |
|---|---|---|---|---|
| 91VIP API | qwen/qwen3.5-122b-a10b | Free | Free | — |
| 91VIP | qwen3.5-122b-a10b | $0.0005 | $0.0011 | — |
| Futureppo | qwen3.5-122b-a10b | $0.0005 | $0.0011 | — |
| 素墨API | qwen/qwen3.5-122b-a10b | $0.010 | $0.010 | — |
 GLM BigModel Relay | qwen3.5-122b-a10b | $0.040 | $0.320 | — |
 OpenRouter Fans | qwen/qwen3.5-9b | $0.050 | $0.150 | — |
| OpenRouter Fans | qwen/qwen3.5-35b-a3b | $0.163 | $1.30 | — |
| OpenRouter Fans | qwen/qwen3.5-27b | $0.195 | $1.56 | — |
| 钱多多 API | Qwen/Qwen3.5-4B | $0.200 | $0.600 | — |
| Seamee API | Qwen/Qwen3.5-35B-A3B | $0.230 | $1.86 | — |
| Seamee API | Qwen/Qwen3.5-27B | $0.260 | $2.09 | — |
| Seamee API | qwen/qwen3.5-122b-a10b | $0.260 | $2.08 | — |
| OpenRouter Fans | qwen/qwen3.5-122b-a10b | $0.260 | $2.08 | — |
| MapleLeaf API | qwen/qwen3.5-122b-a10b | $0.260 | $2.08 | — |
| GPTPlus5 API | qwen3.5-27b | $0.280 | $2.24 | — |
| Seamee API | Qwen/Qwen3.5-122B-A10B | $0.290 | $2.32 | — |
| 人人 API | qwen3.5-122b-a10b | $0.300 | $0.500 | — |
| GPTPlus5 API | qwen3.5-122b-a10b | $0.370 | $2.96 | — |
| SWT-API | qwen3.5-122b-a10b | $0.375 | $0.375 | — |
| SWT-API | qwen3.5-4b | $0.375 | $0.375 | — |
Pricing data from provider public APIs
Qwen3 5 is free to use through 1 provider with no per-token charges. These providers offer free API credits or a free tier:
| Provider | Speed (t/s) |
|---|---|
| 91VIP API | — |
Compare speed and latency performance across all API providers.
Showing 1-7 of 7 providers
| Provider | Speed | Latency | Tests |
|---|---|---|---|
Koru API qwen3.5-9b | 164.85 tok/s | 6.81s | 5 |
 Hugging FaceQwen/Qwen3.5-9B | 108.96 tok/s | 7.99s | 5 |
 SiliconFlowQwen/Qwen3.5-4B | 91.28 tok/s | 23.54s | 5 |
Koru API qwen3.5-27b | 63.56 tok/s | 13.78s | 5 |
 GPT Load (PP.UA)qwen/qwen3.5-122b-a10b | 60.74 tok/s | 0.78s | 15 |
 NVIDIA NIMqwen/qwen3.5-122b-a10b | 42.97 tok/s | 2.80s | 70 |
| SiliconFlow Qwen/Qwen3.5-27B | 34.21 tok/s | 33.63s | 5 |
Latest benchmark results measuring API response speed and first-token latency.
| Time | Model | Speed | Latency |
|---|---|---|---|
| 04/15/2026, 18:51 | qwen/qwen3.5-122b-a10b | 87.95 tok/s | 0.39s |
| 04/15/2026, 18:51 | qwen/qwen3.5-122b-a10b | 128.25 tok/s | 0.25s |
| 04/15/2026, 18:51 | qwen/qwen3.5-122b-a10b | 125.18 tok/s | 0.27s |
| 04/15/2026, 18:51 | qwen/qwen3.5-122b-a10b | 43.68 tok/s | 0.26s |
| 04/15/2026, 18:51 | qwen/qwen3.5-122b-a10b | 49.71 tok/s | 0.26s |
| 04/05/2026, 14:01 | Qwen/Qwen3.5-27B | 47.62 tok/s | 39.90s |
| 04/05/2026, 14:01 | Qwen/Qwen3.5-27B | 30.87 tok/s | 33.48s |
| 04/05/2026, 14:01 | Qwen/Qwen3.5-27B | 44.57 tok/s | 22.87s |
| 04/05/2026, 14:01 | Qwen/Qwen3.5-27B | 3.47 tok/s | 50.97s |
| 04/05/2026, 14:01 | Qwen/Qwen3.5-27B | 44.53 tok/s | 20.95s |
| 04/02/2026, 24:41 | qwen/qwen3.5-122b-a10b | 60.75 tok/s | 0.86s |
| 04/02/2026, 24:41 | qwen/qwen3.5-122b-a10b | 23.56 tok/s | 3.05s |
| 04/02/2026, 24:41 | qwen/qwen3.5-122b-a10b | 58.51 tok/s | 0.79s |
| 04/02/2026, 24:41 | qwen/qwen3.5-122b-a10b | 55.37 tok/s | 0.34s |
| 04/02/2026, 24:41 | qwen/qwen3.5-122b-a10b | 43.24 tok/s | 1.47s |
| 04/01/2026, 08:35 | qwen/qwen3.5-122b-a10b | 15.42 tok/s | 0.62s |
| 04/01/2026, 08:35 | qwen/qwen3.5-122b-a10b | 12.48 tok/s | 2.85s |
| 04/01/2026, 08:35 | qwen/qwen3.5-122b-a10b | 15.85 tok/s | 0.29s |
| 04/01/2026, 08:35 | qwen/qwen3.5-122b-a10b | 67.16 tok/s | 0.58s |
| 04/01/2026, 08:35 | qwen/qwen3.5-122b-a10b | 21.59 tok/s | 12.66s |
