deepseek-v3-1
Developer: DeepSeek
Also known as
DeepSeek V3.1 by DeepSeek is available through 83 API providers on LMSpeed. Compare API pricing from $0.0010 to $200.00 per million input tokens across providers. Free API access is offered by 3 providers. In speed benchmarks, the fastest provider reaches 258 tok/s.
Compare DeepSeek V3.1 API pricing across 80 providers. Input prices range from $0.0010 to $200.00 per million input. 10dian-API offers the lowest rate at $0.0010/M.
| Provider | Model Variant | Input ($/M) | Output ($/M) | Speed (t/s) |
|---|---|---|---|---|
| deepseek-ai/deepseek-v3.1 | Free | Free | — | |
| 91VIP API | deepseek-ai/deepseek-v3.1 | Free | Free | — |
| MapleLeaf API | deepseek-ai/deepseek-v3.1 | Free | Free | — |
| 10dian-API | deepseek-v3.1 | $0.0010 | $0.0010 | — |
| 素墨API | deepseek-ai/deepseek-v3.1 | $0.010 | $0.010 | — |
| 素墨API | deepseek-v3.1 | $0.010 | $0.010 | — |
| CookingAI | deepseek-v3.1 | $0.010 | $0.010 | — |
| 紫脑喵 | deepseek-v3.1 | $0.010 | $0.010 | — |
| uglycat | deepseek-ai/deepseek-v3.1 | $0.050 | $0.050 | — |
 GLM BigModel Relay | deepseek-ai/DeepSeek-V3.1 | $0.056 | $0.168 | — |
 星见雅 API | 李/deepseek-ai/deepseek-v3.1 | $0.100 | $0.100 | — |
| Thorbase | deepseek/deepseek-v3.1 | $0.200 | $0.800 | — |
| 人人 API | deepseek-v3.1 | $0.300 | $0.500 | — |
 Meta API | deepseek-ai/DeepSeek-V3.1 | $0.372 | $1.49 | — |
| SWT-API | deepseek-v3.1 | $0.375 | $0.375 | — |
 UniAiX | deepseek-v3.1 | $0.548 | $1.64 | — |
| EasyMore | deepseek-v3.1 | $0.548 | $1.64 | — |
 天絮 API | DeepSeek-V3.1 | $0.560 | $1.68 | — |
| 天絮 API | deepseek-v3.1 | $0.560 | $1.68 | — |
 352287 API | deepseek-v3.1 | $0.560 | $1.68 | — |
Pricing data from provider public APIs
DeepSeek V3.1 is free to use through 3 providers with no per-token charges. These providers offer free API credits or a free tier:
| Provider | Speed (t/s) |
|---|---|
| — | |
| 91VIP API | — |
| MapleLeaf API | — |
Compare speed and latency performance across all API providers.
Showing 1-10 of 10 providers
| Provider | Speed | Latency | Tests |
|---|---|---|---|
GPT Load (AllAI) DeepSeek-V3.1 | 257.63 tok/s | 0.55s | 5 |
 SophNetDeepSeek-V3.1-Fast | 146.54 tok/s | 0.40s | 15 |
DeepSeek-V3.1-Fast:4T1MSlXFTwTvL7SxmbKNM2 | 93.77 tok/s | 2.01s | 5 |
deepseek-v3.1:671b | 72.05 tok/s | 2.25s | 10 |
 deepseek-ai/DeepSeek-V3.1 | 67.78 tok/s | 3.02s | 31 |
 starlight/DeepSeek-V3.1 | 60.83 tok/s | 1.15s | 5 |
 deepseek-v3-1-250821 | 59.10 tok/s | 1.39s | 30 |
 deepseek-v3.1 | 54.72 tok/s | 1.89s | 40 |
 deepseek-ai/deepseek-v3.1 | 41.41 tok/s | 2.36s | 30 |
 DeepSeek-V3.1-free | 26.67 tok/s | 2.51s | 5 |
Latest benchmark results measuring API response speed and first-token latency.
| Time | Model | Speed | Latency |
|---|---|---|---|
| 12/16/2025, 06:27 | deepseek-v3.1 | 114.30 tok/s | 0.92s |
| 12/16/2025, 06:27 | deepseek-v3.1 | 110.42 tok/s | 0.40s |
| 12/16/2025, 06:27 | deepseek-v3.1 | 154.59 tok/s | 0.41s |
| 12/16/2025, 06:27 | deepseek-v3.1 | 145.93 tok/s | 0.44s |
| 12/16/2025, 06:27 | deepseek-v3.1 | 136.26 tok/s | 0.58s |
| 11/25/2025, 13:11 | deepseek-v3.1:671b | 75.88 tok/s | 7.38s |
| 11/25/2025, 13:11 | deepseek-v3.1:671b | 67.41 tok/s | 0.90s |
| 11/25/2025, 13:11 | deepseek-v3.1:671b | 82.11 tok/s | 0.94s |
| 11/25/2025, 13:11 | deepseek-v3.1:671b | 56.35 tok/s | 0.86s |
| 11/25/2025, 13:11 | deepseek-v3.1:671b | 78.31 tok/s | 3.09s |
| 11/25/2025, 12:02 | deepseek-v3.1:671b | 74.70 tok/s | 2.26s |
| 11/25/2025, 12:02 | deepseek-v3.1:671b | 79.61 tok/s | 1.98s |
| 11/25/2025, 12:02 | deepseek-v3.1:671b | 45.54 tok/s | 1.45s |
| 11/25/2025, 12:02 | deepseek-v3.1:671b | 75.37 tok/s | 1.96s |
| 11/25/2025, 12:02 | deepseek-v3.1:671b | 85.20 tok/s | 1.74s |
| 11/06/2025, 15:16 | deepseek-ai/DeepSeek-V3.1 | 57.34 tok/s | 4.40s |
| 11/06/2025, 15:16 | deepseek-ai/DeepSeek-V3.1 | 38.48 tok/s | 3.01s |
| 11/06/2025, 15:16 | deepseek-ai/DeepSeek-V3.1 | 54.66 tok/s | 11.70s |
| 11/06/2025, 15:16 | deepseek-ai/DeepSeek-V3.1 | 36.95 tok/s | 2.96s |
| 11/06/2025, 15:16 | deepseek-ai/DeepSeek-V3.1 | 64.60 tok/s | 3.53s |
