Kimi K2.5
Open weightVisionToolsMoonshotAI · Released January 2026
Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...
It belongs to Moonshot AI's Kimi K2 family, open-weight models known for agentic tool use, with Code and Thinking variants for their namesake strengths.
Facts and pricing
- Provider
- MoonshotAI
- Context window
- 262K tokens
- Input price
- 4.1 credits / 1M tokens ($0.38 raw)
- Output price
- 22.3 credits / 1M tokens ($2.02 raw)
- Vision (image input)
- Yes
- Tool calling
- Yes
- Extended reasoning
- No
Credits are what Zeplik bills: 1 credit = $0.10, computed from the raw provider rate with a 1.10x margin. Raw prices shown per 1M tokens.
Try Kimi K2.5 now
What Kimi K2.5 is best for
- Everyday chat and drafting on an open-weight model with transparent lineage
- Long documents: the 262K-token window holds lengthy reports, contracts or papers whole
- Working with images: screenshots, charts, photos and scanned documents alongside text
- Tool use and agents: reliably calls functions, so it can search, run skills and drive workflows
Example prompts
Prompts that suit a open weight model like Kimi K2.5:
- Draft a project update from these rough notes
- Explain how HTTPS works to a curious teenager
- Turn this list of features into a changelog entry
- Brainstorm objections to this proposal and how to answer them
MoonshotAI family
| Model | Context | Input cr/M | Output cr/M | Released |
|---|---|---|---|---|
| Kimi K2.7 Code | 262K | 7.9 | 38.5 | June 2026 |
| Kimi K2.6 | 262K | 7.2 | 37.5 | April 2026 |
| Kimi K2.5 | 262K | 4.1 | 22.3 | January 2026 |
| Kimi K2 Thinking | 262K | 6.6 | 27.5 | November 2025 |
| Kimi K2 0905 | 262K | 6.6 | 27.5 | September 2025 |
| Kimi K2 0711 | 131K | 6.3 | 25.3 | July 2025 |
Frequently asked questions
- What is Kimi K2.5?
- Kimi K2.5 is an AI model by MoonshotAI. Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...
- How much does Kimi K2.5 cost on Zeplik?
- Input tokens cost 4.1 credits per million and output tokens 22.3 credits per million (1 credit = $0.10; the raw provider rates are $0.38 and $2.02 per million). New accounts start with free credits.
- How long can a conversation with Kimi K2.5 be?
- Kimi K2.5 has a 262K-token context window (262,144 tokens), which covers the conversation plus any documents you attach.
- Does Kimi K2.5 support images and tools?
- Kimi K2.5 accepts image input and supports tool calling.
Related models
MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long contexts. It uses a native multimodal mixture-of-experts...
Kimi K2.6 is Moonshot AI's next-generation multimodal model, designed for long-horizon coding, coding-driven UI/UX generation, and multi-agent orchestration. It handles complex end-to-end coding tasks across Python, Rust, and Go, and...
Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...
Kimi K2 0905 is the September update of [Kimi K2 0711](moonshotai/kimi-k2). It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32...
Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It is optimized for...
GPT-5.6 Luna Pro is the same underlying model as [GPT-5.6 Luna](https://openrouter.ai/openai/gpt-5.6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode