Chinese AI start-up Moonshot has unveiled the Kimi K3 large language model boasting one million context tokens. The model is built to rival Anthropic’s Opus 4.8 per competitive coding benchmarks found in early tests. Benchmark leaks show Kimi K3 performance targets Anthropic’s Claude Fable5 openly. Industry watchers argue the release escalates the AI competition from Chinese and Western developers. Moonshot frames the model as a step in narrowing the performance gap with leading global alternatives.