Kimi K2.6: benchmarks, price and rank
Kimi K2.6 from Moonshot is 54th of 123 rated AI models on Spezia's leaderboard, strongest on LiveCodeBench v6. It costs $0.75 in, $3.50 out per million tokens.
- 54th of 123 rated models
- Rating 1,301
- Rated on 15 benchmarks
- Open weights
- Released Apr 20, 2026
- 262,000 tokens of context
Where it is strongest
- 5th of 43 on LiveCodeBench v6
Where it falls behind
- 54th of 78 on SimpleQA Verified
- 45th of 66 on FrontierMath Tier 4
- 5th of 8 on DeepSearchQA
Price: $0.75 in, $3.50 out per million tokens.
Its headline results
The benchmarks that count most in the rating, with its place among the models measured the same way, and where each number came from.
- APEX-Agents27.9%
Long tasks written by investment banking analysts, management consultants and corporate lawyers from their own work, done across files and apps in a realistic work setup.
The maker's own number, from Moonshot AI, via LLM Stats.
- BrowseComp86.3%
Questions whose answers are hidden deep on the web and take many searches and pages to track down.
8th of 56. The maker's own number, from Moonshot AI, via LLM Stats.
- SWE-Bench Pro58.6%
Long changes to real codebases in several languages, often across many files, including code from private and copyleft projects that models are unlikely to have seen.
22nd of 50. The maker's own number, from Moonshot AI, via LLM Stats.
- Humanity's Last Exam36.4%
About 2,500 questions written by experts in more than a hundred fields to be as hard as they could make them, many needing a diagram or image.
34th of 74. The maker's own number, from Moonshot AI, via LLM Stats.
- Vending-Bench 2$6,205
The model runs a simulated vending machine business for a year on its own: finding suppliers, ordering stock, setting prices and paying the daily fees.
23rd of 60. Independent result from Epoch AI.
- OSWorld-Verified73.1%
Ordinary tasks on a real desktop computer using normal apps: a browser, a spreadsheet, an email client, a photo editor, the file manager.
15th of 25. The maker's own number, from Moonshot AI, via LLM Stats.
- Toolathlon50%
Long tasks across many real apps and their tools, such as email, calendars, spreadsheets, code repositories and online stores.
19th of 37. The maker's own number, from Moonshot AI, via LLM Stats.
- LiveCodeBench v689.6%
Programming contest problems collected after models were trained, from LeetCode, AtCoder and Codeforces.
5th of 43. The maker's own number, from Moonshot AI, via LLM Stats.
- SimpleQA Verified34.9%
Short factual questions with one checked answer, asked without web search.
54th of 78. Independent result from Epoch AI.
- FrontierMath57.2%
Hundreds of new, unpublished math problems written by mathematicians, from hard undergraduate level up to research, with answers a computer can check.
47th of 76. Independent result from Epoch AI.
12 more results in its full profile.
Every result with its source, what it costs per task, what jumped with its release, side by side with any other model, and answers to your questions about when to use it.
Models ranked near it
How the rating works, and every benchmark it uses: the model leaderboard.