Kimi K2 Thinking
Kimi K2 Thinking is a large language model made by MoonshotAI, added to the catalog on 6 Nov 2025. It has a context window of 262,144 tokens and costs $0.60 per 1M input tokens and $2.50 per 1M output tokens. It accepts text and produces text. Typical uses: autonomous research across diverse subjects, dynamic coding assistance for developers and writing support for long-form content creation.
Price checked · source: OpenRouter
What Kimi K2 Thinking is
Kimi K2 Thinking, released by MoonshotAI on November 6, 2025, is an advanced reasoning model designed for complex tasks. It features a massive context window of 256k tokens, allowing it to handle extensive input and output in text format. This model excels in persistent, step-by-step reasoning, making it ideal for tasks that require dynamic tool invocation and complex workflows. With its ability to maintain stable multi-agent behavior through numerous tool calls, Kimi K2 Thinking is optimized for autonomous research, coding, and writing, supporting a wide range of applications in analytical tasks.
Key features
- Trillion-parameter Mixture-of-Experts architecture
- 256k-token context window for extensive tasks
- Supports step-by-step reasoning and tool use
- Optimized for complex workflows and multi-agent behavior
- High inference efficiency for demanding tasks
Typical uses
- Autonomous research across diverse subjects
- Dynamic coding assistance for developers
- Writing support for long-form content creation
- Complex analytical task management
- Real-time tool invocation for enhanced productivity
Kimi K2 Thinking pricing
US dollars, list price through OpenRouter.
| Item | Standard |
|---|---|
| Input, per 1M tokens | $0.60 |
| Output, per 1M tokens | $2.50 |
| Cached input read, per 1M tokens | $0.15 |
No price change since it was added on 6 Nov 2025.
| What a task costs | Cost |
|---|---|
| Answer a short question500 in / 300 out | $0.0011 |
| Summarise a 50-page PDF25,000 in / 800 out | $0.017 |
| Review a 2,000-line source file30,000 in / 2,000 out | $0.023 |
| Read a 300-page contract set150,000 in / 1,500 out | $0.094 |
Kimi K2 Thinking specifications
| Model id | moonshotai/kimi-k2-thinking |
|---|---|
| Version | kimi-k2-thinking-20251106 |
| Context window | 262,144 tokensAbout 196,608 English words |
| Max output | 98,304 tokens |
| Knowledge cutoff | Not published |
| Added to catalog | 6 Nov 2025 |
| Input types | text |
|---|---|
| Output types | text |
| Tool calling | Yes |
| Structured outputs | Yes |
| Reasoning effort control | No |
| Content moderation | Not applied by provider |
Sample outputs
The same prompts are given to every model, and the unedited output is shown with the date it was produced. Upend does not publish leaderboard scores it did not run.
Write a short essay
RUN20 Nov 2026Write 150 words on how AI might positively upend work, leisure and creativity
AI is poised to transform our lives positively across three domains. At work, it will automate repetitive tasks, freeing us for strategic thinking and interpersonal connection. Rather than replacing humans, AI will augment our capabilities, making expertise more accessible and enabling smaller teams to achieve ambitious goals. For leisure, AI will curate hyper-personalized experiences—from entertainment to travel—while creating entirely new forms of interactive recreation. The time saved from mundane tasks becomes space for deeper relaxation and meaningful pursuits. In creativity, AI democratizes expression: a writer can visualize scenes instantly, a musician can compose with virtual orchestras, and anyone can prototype ideas regardless of technical skill. It becomes a collaborative partner that suggests possibilities, breaks through blocks, and handles execution, allowing creators to focus on vision and emotion. This symbiosis between human imagination and machine capability promises not just efficiency, but an expansion of what we can dream and achieve.
Compare Kimi K2 Thinking
Side by side with other models: prices, context window, what each accepts and supports, and the same four tasks priced on both.