Moonshot AI released the open-source Kimi K2.5 model in January 2026, featuring a Mixture-of-Experts architecture with 1 trillion total parameters and only 32 billion activated per request. The model introduces Agent Swarm technology, allowing it to coordinate up to 100 specialized AI agents simultaneously for parallel task execution.

  • Kimi K2.5 achieves 50.2% on Humanity’s Last Exam at 76% lower cost than Claude Opus 4.5.
  • The model operates in four modes: Instant, Thinking, Agent, and Agent Swarm.
  • Agent Swarm cuts execution time by 4.5x compared to sequential processing.
  • Training utilized 15 trillion tokens of mixed visual and textual data.
  • Native INT4 quantization provides 2x speed improvements without accuracy degradation.

The parallel approach enables efficient local deployment while maintaining frontier capabilities for coding, vision, and mathematical reasoning tasks.