Kimi K2 Thinking: API pricing, providers and specs
Kimi K2 Thinking is a text model by Moonshot AI. 10 providers serve it, from $1.68 to $2.70 /1M; the lowest price is at APIMart.
How Kimi K2 Thinking compares
Against the 737 text models on the site priced the same way.
- Mid-priced: its lowest price is $1.68 /1M against a median of $1.10; 60% of comparable models cost less.
- Its 262K-token context window is larger than 45% of them (median 262K), about 390 pages of text.
- 10 providers serve it; 77% of comparable models have fewer.
- Its weights are open under the modified-mit license: commercial use under the maker's terms.
Kimi K2 Thinking prices by provider
Every provider's list price, read from its own price page or API. A price marked “from” changes with context length, resolution or mode.
| Provider | Input | Output | Prompt caching | Context |
|---|---|---|---|---|
| APIMart | $0.40/1M | $1.68/1M | — | — |
 Helicone | $0.48/1M | From$2/1M | — | — |
 AIHubMix | $0.548/1M | $2.19/1M | $0.137/1M | 262K |
 Novita AI | $0.60/1M | $2.50/1M | $0.15/1M | — |
 Charm Hyper | $0.60/1M | $2.50/1M | $0.30/1M | — |
| OpenRouter | $0.60/1M | From$2.50/1M | $0.15/1M | 262K |
 Azure | $0.60/1M | From$2.50/1M | — | — |
 Amazon Bedrock | $0.60/1M | From$2.50/1M | — | — |
| AI/ML API | $0.635/1M | $2.65/1M | $0.159/1M | — |
| Replicate | $0.66/1M | $2.70/1M | — | — |
- APIMart$1.68
- HeliconeFrom$2
- AIHubMix$2.19
- Novita AI$2.50
- Charm Hyper$2.50
- OpenRouterFrom$2.50
- AzureFrom$2.50
- Amazon BedrockFrom$2.50
- AI/ML API$2.65
- Replicate$2.70
Specs
- Type
- Text
- Maker
- Moonshot AI
- Inputs
- Text
- Outputs
- Text
- Context
- 262K≈ 390 pages of text
- Max output
- 98K
- Capabilities
- Prompt caching, JSON, Reasoning, Tool calling
- Released
- Nov 4, 2025
- License
- modified-mitcommercial use under the maker's terms
- Open weights
- Yes, downloadable
- Parameters
- 1.03T
Alternatives to Kimi K2 Thinking
Same kind of model from other makers, at a similar price.
| Model | Lowest price | Context | Arena | Providers |
|---|---|---|---|---|
| Kimi K2 Thinkingthis model | $1.68/1M | 262K | — | 10 |
 GLM-5.3 | $1.10/1M | 1.05M | 1623 | 49 |
 Qwen3.8 2.4T A95B | $4.11/1M | 1M | — | 22 |
| Grok 4.7 | $4.80/1M | 500K | 1637 | 28 |
 Gemini 3 Flash | $2.40/1M | 1.05M | 1466 | 5 |
 Mistral Large 4 | $2.09/1M | 1.05M | 1543 | 13 |
 Muse Spark 1.3 | $4.25/1M | 1.05M | 1657 | 11 |
 Sonar | $1/1M | 127K | — | 9 |
| GLM-5.2 | $1.80/1M | 1M | 1604 | 44 |
More from Moonshot AI
Kimi K2 Thinking FAQ
The lowest price for Kimi K2 Thinking is $1.68 /1M, at APIMart.
APIMart, at $1.68 /1M. 10 providers serve Kimi K2 Thinking; the table above lists every one with its price.
Kimi K2 Thinking reads up to 262K tokens per request and writes up to 98K.
Yes. Kimi K2 Thinking's weights can be downloaded from its maker's Hugging Face page, under the modified-mit license.
About 1.03T, as the safetensors metadata of its official Hugging Face repo states.
10 API providers serve Kimi K2 Thinking. Compare their prices above, or open a provider to see everything it offers.
Can't find what you're looking for? Contact us