# Kimi K2 Thinking

> Kimi K2 Thinking is an open-weight language model from Moonshot AI, released on 6 Nov 2025. It accepts text and generates text, with a 262K-token context window and up to 236K output tokens per response. As of 4 Oct 2026, the lowest listed API price is $0.47 per million input tokens and $2 per million output tokens (Vercel AI Gateway), across 5 providers we track.

Source page: https://www.aimodel.directory/models/kimi-k2-thinking
Last verified: 4 Oct 2026

## Specifications

- **Developer:** Moonshot AI
- **Release date:** 6 Nov 2025 (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Status:** Generally available (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Weights:** Open weights (source: Hugging Face, https://huggingface.co/moonshotai/Kimi-K2-Thinking)
- **License:** modified-mit (source: Hugging Face, https://huggingface.co/moonshotai/Kimi-K2-Thinking)
- **Parameters:** 1T (source: Hugging Face, https://huggingface.co/moonshotai/Kimi-K2-Thinking)
- **Context window:** 262K (262,144 tokens) (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Max output:** 236K tokens (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Knowledge cutoff:** Aug 2024 (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Input:** Text (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Output:** Text (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Reasoning mode:** Yes (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Tool calling:** Yes (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Structured output:** Yes (source: models.dev, https://github.com/sst/models.dev/blob/dev/providers/openrouter/models/moonshotai/kimi-k2-thinking.toml)
- **Hugging Face:** moonshotai/Kimi-K2-Thinking (source: Hugging Face, https://huggingface.co/moonshotai/Kimi-K2-Thinking)

## API pricing (USD per 1M tokens)

| Provider | Input | Output | Cached input | Context |
|---|---|---|---|---|
| Vercel AI Gateway | $0.47 | $2 | $0.14 | 216K |
| Amazon Bedrock | $0.60 | $2.50 | – | 262K |
| Novita AI | $0.60 | $2.50 | $0.15 | 262K |
| Hugging Face Inference Providers | $0.60 | $2.50 | $0.15 | 262K |
| OpenRouter | $0.60 | $2.50 | – | 262K |

## FAQ

### What is the context window of Kimi K2 Thinking?

Kimi K2 Thinking has a context window of 262,144 tokens (262K) and can generate up to 235,929 tokens in a single response, according to models.dev.

### How much does the Kimi K2 Thinking API cost?

Kimi K2 Thinking is listed from $0.47 per million input tokens and $2 per million output tokens (Vercel AI Gateway). Prices last verified 4 Oct 2026.

### Is Kimi K2 Thinking open source?

Kimi K2 Thinking is an open-weight model: its weights are published on Hugging Face as moonshotai/Kimi-K2-Thinking under the modified-mit license. Check the license terms before commercial use.

### When was Kimi K2 Thinking released?

Moonshot AI released Kimi K2 Thinking on 6 Nov 2025, according to models.dev.

### Which providers offer Kimi K2 Thinking?

We track Kimi K2 Thinking on 5 providers: Vercel AI Gateway, Amazon Bedrock, Novita AI, Hugging Face Inference Providers, OpenRouter.

---
Data licensed CC BY 4.0. Cite as: AI Model Directory, "Kimi K2 Thinking", https://www.aimodel.directory/models/kimi-k2-thinking