Moonshot AI Launches Kimi K3: 2.8T Parameters, 1M-Token Context Window

Moonshot AI Launches Kimi K3: 2.8T Parameters, 1M-Token Context Window

Moonshot AI launched Kimi K3, a 2.8-trillion-parameter, natively multimodal model with a 1-million-token context window — roughly 750,000 words per inference call. The scale matters: that context length can absorb entire codebases or legal corpora without chunking. No independent benchmarks yet exist; specs are from Moonshot's own disclosures.

Published

Read at another depth