Moonshot AIopen-weight
Released: 2026-07-26Kimi K3
Moonshot AI massive 2.8 trillion parameter open-weight MoE model featuring Kimi Delta Attention (KDA) and 1M context reasoning.
Intelligence Index
93.5/ 100
Calibrated multi-domain compositeThroughput Speed
58tok/sec
Output streaming throughputTime To First Token
460ms
Initial chunk server latencyLiveBench Contam-Free
82.0%
September 2026 suite scoreLiveBench Multi-Domain Evaluation Breakdown
Monthly refreshed contamination-free test setsLogical Reasoning90.7%
Coding & Repo Repair83.5%
Mathematics (AIME)87.0%
Data Analysis & Tables79.0%
Language & Comprehension76.0%
Instruction Following76.0%
Sourced Benchmark Evaluations (3)
Verified performance across authoritative benchmarks with provenance tracking.Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run kimi-k3
# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "kimi-k3",
"messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
}'Evaluation & Sourcing Notes
Scores listed for Kimi K3 represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.