Kimi K2.5
Kimi K2.5 is a multimodal model with a 262,144-token context window. It is closed to new API users and the platform sunsets it on August 31, 2026.
Last updated
kimi-k2.5Key facts
At a glance
- Context window
- 262,144
- Parameters
- 1T
- Released
- January 1, 2026
- Output / 1M
- $3.00
Pricing
How much does Kimi K2.5 cost?
Kimi K2.5 costs $0.60 per million input tokens on a cache miss, $0.10 per million on a cache hit, and $3.00 per million output tokens. Prices exclude tax.
| Token type | Price per 1M tokens |
|---|---|
| Input — cache hit | $0.10 |
| Input — cache miss | $0.60 |
| Output | $3.00 |
What a real request costs
A 100,000-token prompt with a 5,000-token response costs $0.075 cold. Repeat that same prompt prefix so 80% of the input hits the cache and it drops to $0.035 — a 53% saving.
This is why headline per-token prices mislead on agent workloads, where the system prompt and tool definitions repeat on every call.
Capabilities
What can Kimi K2.5 do?
- Text inputsupported
- Image inputsupported
- Video inputsupported
- Thinking modesupported
- Tool callingsupported
- JSON modesupported
- Structured outputnot supported
- Partial modesupported
- Context cachingsupported
- Web searchsupported
- Open weightssupported
Notes
What the spec sheet leaves out
- Not available to accounts registered after the Kimi K3 launch.
- Full platform sunset is scheduled for August 31, 2026 — existing users should migrate.
- Still the cheapest Kimi model per token while it remains available.
- Moonshot has published the month but not an exact release day; January 2026 is what we can confirm.
FAQ
Kimi K2.5 questions
What is Kimi K2.5?
Kimi K2.5 is a multimodal model with a 262,144-token context window. It is closed to new API users and the platform sunsets it on August 31, 2026.
What is the Kimi K2.5 context window?
Kimi K2.5 has a 262,144-token context window. That total covers input and output combined, so a long prompt reduces the room left for the response.
How much does Kimi K2.5 cost?
Kimi K2.5 costs $0.60 per million input tokens on a cache miss, $0.10 per million on a cache hit, and $3.00 per million output tokens. Prices exclude tax.
When does Kimi K2.5 shut down?
Moonshot AI sunsets Kimi K2.5 on August 31, 2026. It is already closed to newly registered accounts, so plan a migration to Kimi K3 or Kimi K2.7 Code before that date.
Is Kimi K2.5 open source?
Moonshot AI published downloadable open weights for Kimi K2.5, so you can self-host it. That is not the same as an OSI-approved open-source licence — check the licence attached to the weight release before deploying commercially.
Sources
Where these figures come from
Every specification and price on this page is transcribed from Moonshot AI’s own documentation and re-checked on the date shown. Nothing is estimated.