kimi-k2-thinking
Developer: Moonshot AI
Also known as
Kimi K2 Thinking by Moonshot AI is available through 64 API providers on LMSpeed. Compare API pricing from $0.0009 to $360.00 per million input tokens across providers. Free API access is offered by 2 providers. In speed benchmarks, the fastest provider reaches 131 tok/s.
Compare Kimi K2 Thinking API pricing across 62 providers. Input prices range from $0.0009 to $360.00 per million input. 91VIP offers the lowest rate at $0.0009/M.
| Provider | Model Variant | Input ($/M) | Output ($/M) | Speed (t/s) |
|---|---|---|---|---|
| moonshotai/kimi-k2-thinking | Free | Free | — | |
| 91VIP API | moonshotai/kimi-k2-thinking | Free | Free | — |
| 91VIP | kimi-k2-thinking | $0.0009 | $0.0019 | — |
| Futureppo | kimi-k2-thinking | $0.0009 | $0.0019 | — |
| Medu Chat | kimi-k2-thinking | $0.0010 | $0.0010 | — |
| 素墨API | kimi-k2-thinking | $0.010 | $0.010 | — |
| 素墨API | moonshotai/kimi-k2-thinking | $0.010 | $0.010 | — |
| CookingAI | kimi-k2-thinking | $0.010 | $0.010 | — |
| 紫脑喵 | kimi-k2-thinking | $0.010 | $0.010 | — |
 GLM BigModel Relay | kimi-k2-thinking | $0.040 | $0.250 | — |
 WONG公益站 | moonshotai/kimi-k2-thinking | $0.300 | $1.20 | — |
| Seamee API | moonshotai/kimi-k2-thinking | $0.300 | $1.20 | — |
 OpenRouter Fans | moonshotai/kimi-k2-thinking | $0.300 | $1.20 | — |
| MapleLeaf API | moonshotai/kimi-k2-thinking | $0.300 | $1.20 | — |
| 人人 API | kimi-k2-thinking | $0.350 | $0.650 | — |
| SWT-API | kimi-k2-thinking | $0.375 | $0.375 | — |
| MapleLeaf API | kimi-k2-thinking | $0.400 | $2.50 | — |
| OpenRouter Fans | moonshotai/Kimi-K2-Thinking | $0.470 | $2.00 | — |
 UniAiX | kimi-k2-thinking | $0.548 | $2.19 | — |
| EasyMore | kimi-k2-thinking | $0.548 | $2.19 | — |
Pricing data from provider public APIs
Kimi K2 Thinking is free to use through 2 providers with no per-token charges. These providers offer free API credits or a free tier:
Compare speed and latency performance across all API providers.
Showing 1-5 of 5 providers
| Provider | Speed | Latency | Tests |
|---|---|---|---|
QYES AI kimi-k2-thinking | 131.31 tok/s | 4.75s | 5 |
accounts/fireworks/models/kimi-k2-thinking | 127.75 tok/s | 3.47s | 5 |
 kimi-k2-thinking | 45.35 tok/s | 7.51s | 5 |
 moonshotai/kimi-k2-thinking | 44.80 tok/s | 15.74s | 50 |
 kimi-k2-thinking | 33.80 tok/s | 16.24s | 10 |
Latest benchmark results measuring API response speed and first-token latency.
| Time | Model | Speed | Latency |
|---|---|---|---|
| 04/08/2026, 17:11 | moonshotai/kimi-k2-thinking | 23.18 tok/s | 27.88s |
| 04/08/2026, 17:11 | moonshotai/kimi-k2-thinking | 33.55 tok/s | 20.58s |
| 04/08/2026, 17:11 | moonshotai/kimi-k2-thinking | 34.12 tok/s | 9.43s |
| 04/08/2026, 17:11 | moonshotai/kimi-k2-thinking | 42.39 tok/s | 19.70s |
| 04/08/2026, 17:11 | moonshotai/kimi-k2-thinking | 18.80 tok/s | 26.02s |
| 04/08/2026, 16:12 | moonshotai/kimi-k2-thinking | 19.14 tok/s | 70.82s |
| 04/08/2026, 16:12 | moonshotai/kimi-k2-thinking | 40.85 tok/s | 16.88s |
| 04/08/2026, 16:12 | moonshotai/kimi-k2-thinking | 24.32 tok/s | 12.52s |
| 04/08/2026, 16:12 | moonshotai/kimi-k2-thinking | 20.43 tok/s | 9.81s |
| 04/08/2026, 16:12 | moonshotai/kimi-k2-thinking | 24.48 tok/s | 38.83s |
| 04/08/2026, 03:53 | kimi-k2-thinking | 135.09 tok/s | 5.14s |
| 04/08/2026, 03:53 | kimi-k2-thinking | 88.49 tok/s | 3.99s |
| 04/08/2026, 03:53 | kimi-k2-thinking | 160.08 tok/s | 3.97s |
| 04/08/2026, 03:53 | kimi-k2-thinking | 119.80 tok/s | 4.93s |
| 04/08/2026, 03:53 | kimi-k2-thinking | 153.09 tok/s | 5.72s |
| 04/07/2026, 06:08 | moonshotai/kimi-k2-thinking | 64.82 tok/s | 6.35s |
| 04/07/2026, 06:08 | moonshotai/kimi-k2-thinking | 105.06 tok/s | 5.68s |
| 04/07/2026, 06:08 | moonshotai/kimi-k2-thinking | 39.80 tok/s | 7.45s |
| 04/07/2026, 06:08 | moonshotai/kimi-k2-thinking | 55.02 tok/s | 8.05s |
| 04/07/2026, 06:08 | moonshotai/kimi-k2-thinking | 91.37 tok/s | 12.07s |
