Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...
Tested on 29 benchmarks · BenchGecko score 68.9. Top scores: OTIS Mock AIME 2024-2025 (97.2%), Surface Evolver Bench (95.0%), ARC-AGI (94.5%).
DeepSeek V4 Flash scores 68.6 (100% as good) at $0.03/1M input · 96% cheaper
Unusual and adversarial machine learning challenges. Tests robustness of reasoning about edge cases in ML systems.
Abstraction and Reasoning Corpus. Tests fluid intelligence through novel visual pattern recognition puzzles. Core measure of general intelligence.
ARC-AGI 2, harder sequel to ARC. More complex abstract reasoning patterns that test generalization ability beyond training data.
Deceptively simple questions that humans find easy but AI models often get wrong. Tests common sense and reasoning gaps.
Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.
- Typemultimodal
- Context1.0M tokens (~524 books)
- ReleasedJul 2026
- LicenseProprietary
- StatusActive
- Cost / Message~$0.015
Frequently Asked Questions
Related Models
Gemini 3 Pro · Google DeepMindClaude Instant · Anthropicphi-3-medium 14B · MicrosoftDeepSeek V4 Flash · DeepSeekMixtral 8x7B Instruct · Mistral AIKey facts · as of 2026-10-05
- Kimi K3 by moonshotai. BenchGecko score 68.9, rank 40 of 312 scored models (normalized average of public benchmark scores).
- List price $0.67 input · $14.00 output per 1M tokens (as of 2026-10-05).
- Sold by 24 providers (as of 2026-10-05): Relace (fp4) $0.63 in / $13.00 out · InferenceNet (fp4) $0.67 in / $14.00 out · Wafer $0.95 in / $15.00 out · Decart (mxfp4) $1.80 in / $9.00 out · Phala $1.95 in / $9.75 out · and 19 more. Every provider
- Gecko Tests: Who Are You B (Knows who made it) · Tokenizer Tax E (93% more tokens outside English).
How to cite · data as of 2026-10-05
Kimi K3 · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/kimi-k3
Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP