moonshotai
MoonshotAI: Kimi K2 0711
moonshotai/kimi-k2
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.
- Context window: 131,072 tokens
- Input: text
- Output: text
- Pricing: $0.57/M input tokens, $2.3/M output tokens
RuntimeWire coverage
- Periodic Labs details the 1,300-GPU stack behind Neon
- Head to head: OpenAI: GPT-5.6 Luna vs Grok 4.20 0309 Non Reasoning
- Moonshot releases Kimi K3 report, detailing how a 2.8T open model runs
- Google Cloud will retire Moonshot AI's Kimi K2 endpoint in October
- AMD's MI455X delivers 432GB HBM4, enough for trillion‑parameter model weights in one server
View on OpenRouter. Model data sourced from OpenRouter.