Agent may make mistakes and shouldn't be used in production
Kimi K2 is a model with a context length of 128k, featuring powerful code and Agent capabilities based on MoE architecture. It has 1T total parameters with 32B activated parameters. In benchmark performance tests across major categories including general knowledge reasoning, programming, mathematics, and Agent capabilities, the K2 model outperforms other mainstream open-source models.
import { streamText } from 'ai'
const result = streamText({ model: 'moonshotai/kimi-k2', prompt: 'Why is the sky blue?'})Try out Kimi K2 by Moonshot AI. Usage is billed to your team at API rates. Free users get $5 of credits every 30 days, and you are considered a free user if you haven't made a payment.
| Model |
|---|
| 262K | 2.0s | 102tps | $0.95/M | $4.00/M | Read:$0.16/M Write:— | — | — |   | 04/20/2026 | ||||
| 262K | 3.6s | 85tps | $0.50/M | $2.80/M | Read:$0.1/M Write:— | — | — |  +2 | 01/26/2026 | ||||
| 262K | 1.0s | 30tps | $0.60/M | $2.50/M | Read:$0.15/M Write:— | — | — |  | 11/06/2025 | ||||
| 262K | 0.7s | 120tps | $1.15/M | $8.00/M | Read:$0.15/M Write:— | — | — | | 11/06/2025 | ||||
| 256K | $0.60/M | $2.50/M | Read:$0.3/M Write:— | — | — | | 09/05/2025 | ||||||
| 256K | 0.9s | $1.15/M | $8.00/M | Read:$0.15/M Write:— | — | — | | 09/05/2025 |
