Kimi K2.6

Leaderboards

Where Kimi K2.6 sits among the top 20 in each category.

By task

Scored from votes on each kind of prompt alone, on the same 1,000-point scale.

Test results

Every test Kimi K2.6 has a published result on, out of 100%, with its rank among the models that took it.

• GPQA

  #24 / 256

  Graduate-level science questions written to be hard to look up.

  90.5%

  Self-reported

• SWE-Bench Verified

  #13 / 120

  Fixing real bugs in real GitHub projects, checked by the projects' own tests.

  80.2%

  Self-reported

• Humanity's Last Exam

  #46 / 108

  Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions

  36.4%

  Self-reported · In the text score· In the vision score

• MMMU-Pro

  #15 / 73

  College-level questions that need both an image and text to answer.

  80.1%

  Self-reported · In the vision score

• BrowseComp

  #8 / 68

  Finding hard-to-locate facts by browsing the web.

  86.3%

  Self-reported · In the agents score

• LiveCodeBench v6

  #6 / 67

  LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.

  89.6%

  Self-reported

• SWE-Bench Pro

  #24 / 61

  Harder, longer software engineering tasks from real codebases.

  58.6%

  Self-reported · In the coding score

• CharXiv-R

  #13 / 60

  Reading and reasoning about charts from scientific papers.

  86.7%

  Self-reported

• Terminal-Bench 2.0

  #15 / 54

  Getting real work done in a command-line terminal.

  66.7%

  Self-reported

• SWE-bench Multilingual

  #13 / 49

  A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.

  76.7%

  Self-reported

• Toolathlon

  #22 / 42

  Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.

  50.0%

  Self-reported · In the agents score

• LiveBench

  #25 / 38

  LiveBench is a challenging, contamination-limited LLM benchmark that addresses test set contamination by releasing new questions monthly based on recently-released datasets, arXiv papers, news articles, and IMDb movie synopses. It comprises tasks across math, coding, reasoning, language, instruction following, and data analysis with verifiable, objective ground-truth answers.

  72.2%

  Self-reported · In the text score

• MathVision

  #6 / 38

  MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.

  93.2%

  Self-reported

• SciCode

  #7 / 30

  SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.

  52.2%

  Self-reported

• Finance Agent v2

  #13 / 28

  Finance Agent v2 is an agentic financial-analysis benchmark from Vals that evaluates models on real-world finance workflows, measuring their ability to retrieve and reason over financial documents, perform multi-step calculations, and produce accurate analyses.

  44.9%

  Self-reported

• OSWorld-Verified

  #18 / 28

  OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.

  73.1%

  Self-reported

• AIME 2026

  #4 / 26

  All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.

  96.4%

  Self-reported

• IMO-AnswerBench

  #7 / 22

  IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.

  86.0%

  Self-reported

• BabyVision

  #10 / 16

  A benchmark for early-stage visual reasoning and perception on child-like vision tasks.

  68.5%

  Self-reported

• FrontierSWE

  #14 / 16

  FrontierSWE measures whether an agent can complete open-ended technical projects at the scale of hours to tens of hours, spanning systems optimization, large-scale code construction, and applied ML research. Performance is reported as a dominance score, where higher is better.

  27.0%

  Self-reported

• Claw-Eval

  #1 / 14

  Claw-Eval tests real-world agentic task completion across complex multi-step scenarios, evaluating a model's ability to use tools, navigate environments, and complete end-to-end tasks autonomously.

  80.9%

  Self-reported

• WideSearch

  #3 / 14

  WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.

  80.8%

  Self-reported

• HMMT Feb 26

  #5 / 13

  HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.

  92.7%

  Self-reported

• APEX-Agents

  #9 / 11

  APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.

  27.9%

  Self-reported

• DeepSearchQA

  #7 / 10

  DeepSearchQA is a benchmark for evaluating deep search and question-answering capabilities, testing models' ability to perform multi-hop reasoning and information retrieval across complex knowledge domains.

  83.0%

  Self-reported

• MCP-Mark

  #4 / 9

  MCP-Mark evaluates LLMs on their ability to use Model Context Protocol (MCP) tools effectively, testing tool discovery, selection, invocation, and result interpretation across diverse MCP server scenarios.

  55.9%

  Self-reported

• OJBench

  #1 / 9

  OJBench is a competition-level code benchmark designed to assess the competitive-level code reasoning abilities of large language models. It comprises 232 programming competition problems from NOI and ICPC, categorized into Easy, Medium, and Hard difficulty levels. The benchmark evaluates models' ability to solve complex competitive programming challenges using Python and C++.

  60.6%

  Self-reported

• V*

  #1 / 7

  A visual reasoning benchmark evaluating multimodal inference under challenging spatial and grounded tasks.

  96.9%

  Self-reported