Local LLM model page

Kimi K2.5 (32B/1T MoE)

Moonshot AI's agentic flagship. 1T total MoE parameters with 32B active per forward pass. Unmatched long-context reasoning at 256K tokens. Designed for complex agentic tasks and tool use. Model License — check moonshotai.com for commercial terms.

Find the best model for my hardwareBrowse all 183 LLMs
Parameters
32B active (1T total MoE)
Minimum RAM
32 GB
Model size
22 GB
Quantization
Q4_K_M

Can Kimi K2.5 (32B/1T MoE) run locally?

Kimi K2.5 (32B/1T MoE) is best suited for power-user machines with 32 GB RAM. LocalClaw recommends Q4_K_M as the default quantization, with at least 32 GB RAM.

Search term for LM Studio or compatible runtimes: kimi-k2.5-32b

Hugging Face repository: moonshotai/Kimi-K2.5-32B-GGUF

chatcodereasoningpowerquality

Strengths

  • Massive 1T MoE with 32B active
  • 256K context
  • Unmatched reasoning
  • Top-tier quality

Limitations

  • Needs 32GB+ RAM
  • Model license restrictions
  • Complex setup

Best use cases

  • Long document processing
  • Complex reasoning
  • Enterprise AI
  • Research

Benchmarks

Speed: 4/10

Quality: 10/10

Coding: 10/10

Reasoning: 10/10

Technical details

Developer: Moonshot AI

License: Model License

Context window: 262,144 tokens

Architecture: Mixture of Experts — 1T total, 32B active

Released: 2026-01

Similar models

qwen3-32bdeepseek-r1-32bcogito-32b