Kimi K3.1 Pro
Active · kimi-k3.1-pro
- Input / 1M
- $2.00
- Cached input / 1M
- $0.20
- Output / 1M
- $10.00
- Context
- 1,048,576
- Max output
- 1,048,576
- Modalities
- text, image → text
- Released
- 23 Oct 2026
- Status
- Active
- API id
- kimi-k3.1-pro
- Updated
- 28 Oct 2026
Which rate this is
Cache miss / cache hit / output
The platform home lists these rates: cache-miss input, cache-hit input, and output. Cache writes are billed separately at $2 per 1M tokens for a 5-minute TTL and $4 per 1M tokens for a 1-hour TTL. A Highspeed variant, kimi-k3.1-pro-highspeed, is $4 input, $0.20 cached input, and $20 output per 1M tokens. The quickstart sets max_completion_tokens to 131,072 by default and allows it up to 1,048,576, which is the max output recorded here. The docs pricing table renders client-side, so the stored rates are the ones on the platform home. Open weights are promised by 4 November 2026.
Sources
Updated 28 Oct 2026.
Related
No other Moonshot AI models are in the ledger.
Same-tier comparisons
- Kimi K3.1 Pro vs Command R+ 08-2024 · Cohere