Kimi K2.7 Code
MoonshotAiSpecialized
ThinkingTool UseVisionStructured Output
About this model
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...
Performance Tier
Specialized
Kimi K2.7 Code is a specialized model from MoonshotAi : built for a specific domain.
Domain-specific model. Optimized for a particular task such as code generation, image creation, or web search.
Pricing
This model is included in Elosia plansModerate
Moderate cost. A balanced choice for regular use without constant cap watching.
| Type | per 1M tokens |
|---|---|
| Input (prompt) | $0.680 |
| Output (completion) | $3.41 |
| Cache read | $0.144 |
Capabilities
Context Length262K
Max Output Tokens262K
TokenizerOther
Inputtext, image
Outputtext
Release DateJune 12, 2026
Recommended Use Cases
Coding
Strengths
- Code-specialized variant of Kimi K2.6, tuned for long-horizon software engineering and agentic coding
- Frontier-scale open weights — 1T-parameter MoE (32B active) under a permissive Modified MIT license, self-hostable
- ~30% fewer reasoning tokens than Kimi K2.6 at equal-or-better coding scores — lower cost per agentic task
- 256K context with text + image input via the MoonViT vision encoder, uncommon for a code-focused model
- Cost-efficient at $0.61/M input and $3.07/M output versus Western flagship coding models
Limitations
- Code-specialized — Moonshot recommends Kimi K2.6 for writing, analysis and general conversation
- Moonshot publishes only proprietary in-house benchmarks; no independent public-suite scores
- Forced thinking mode (no instant mode) and 1T total parameters make self-hosting and simple calls resource-intensive
Resources
This model may use your data for training
