Reasoning benchmark

# Kagi LLM Benchmark leaderboard

> Kagi LLM Benchmark results for 99 AI models, led by Claude Fable 5 at 91.4%. What the benchmark measures, who runs it, and a source for every score.
- Canonical page: https://noometry.com/benchmarks/kagi-reasoning
- Last updated: 2026-10-10
- Title: Kagi LLM Benchmark Leaderboard (October 2026): Scores by Model

As of October 2026, Claude Fable 5 has the highest published Kagi LLM Benchmark score on Noometry at 91.4%, out of 99 models with results.

Last verified October 10, 2026

## About Kagi LLM Benchmark

A private, frequently rotated set of reasoning, coding and instruction-following tasks from Kagi, designed so models cannot be trained on it.

- **Category:** [Reasoning](https://noometry.com/best/reasoning)
- **Introduced:** 2024
- **Format:** Mixed private tasks
- **Unit:** Percent (random guessing ≈ 0%)
- **Official site:** [help.kagi.com](https://help.kagi.com/kagi/ai/llm-benchmark.html)

## Top 15 models

Top models on Kagi LLM Benchmark

1.  Claude Fable 5 91.4%
2.  Claude Opus 4.8 88.8%
3.  GPT-5.5 88.8%
4.  Claude Opus 4.6 83.6%
5.  Grok 4.5 83.5%
6.  Claude Opus 4.7 80.7%
7.  Claude Opus 4.5 80.2%
8.  Gemini 3 Pro 80.1%
9.  Kimi K2.5 78.5%
10.  GPT-5 Pro 76.8%
11.  Chatgpt 4o Latest 20250326 75%
12.  GLM-5 75%
13.  Grok 4.20 (Non-Reasoning) 75%
14.  Claude Opus 4 74.3%
15.  Qwen3.5 397B-A17B 73.7%
16.  707580859095

Sponsored placements are available on pages like this one. [Advertise on Noometry](https://noometry.com/advertise)

## All results

Kagi LLM Benchmark results by model
| # | Model | Provider | Score | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- | --- |
| 1 | [Claude Fable 5](https://noometry.com/models/claude-fable-5) | [Anthropic](https://noometry.com/providers/anthropic) | 91.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 2 | [Claude Opus 4.8](https://noometry.com/models/claude-opus-4-8) | [Anthropic](https://noometry.com/providers/anthropic) | 88.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 3 | [GPT-5.5](https://noometry.com/models/gpt-5-5) | [OpenAI](https://noometry.com/providers/openai) | 88.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 4 | [Claude Opus 4.6](https://noometry.com/models/claude-opus-4-6) | [Anthropic](https://noometry.com/providers/anthropic) | 83.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 5 | [Grok 4.5](https://noometry.com/models/grok-4-5) | [xAI](https://noometry.com/providers/xai) | 83.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 6 | [Claude Opus 4.7](https://noometry.com/models/claude-opus-4-7) | [Anthropic](https://noometry.com/providers/anthropic) | 80.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 7 | [Claude Opus 4.5](https://noometry.com/models/claude-opus-4-5) | [Anthropic](https://noometry.com/providers/anthropic) | 80.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 8 | [Gemini 3 Pro](https://noometry.com/models/gemini-3-pro) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 80.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 9 | [Kimi K2.5](https://noometry.com/models/kimi-k2-5) | [Moonshot AI](https://noometry.com/providers/moonshot) | 78.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 10 | [GPT-5 Pro](https://noometry.com/models/gpt-5-pro) | [OpenAI](https://noometry.com/providers/openai) | 76.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 11 | [Chatgpt 4o Latest 20250326](https://noometry.com/models/chatgpt-4o) | [OpenAI](https://noometry.com/providers/openai) | 75% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 12 | [GLM-5](https://noometry.com/models/glm-5) | [Z.ai (Zhipu)](https://noometry.com/providers/zai) | 75% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 13 | [Grok 4.20 (Non-Reasoning)](https://noometry.com/models/grok-4-20) | [xAI](https://noometry.com/providers/xai) | 75% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 14 | [Claude Opus 4](https://noometry.com/models/claude-opus-4) | [Anthropic](https://noometry.com/providers/anthropic) | 74.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 15 | [Qwen3.5 397B-A17B](https://noometry.com/models/qwen3-5-397b-a17b) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 73.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 16 | [Grok 4](https://noometry.com/models/grok-4) | [xAI](https://noometry.com/providers/xai) | 73.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 17 | [GPT-5.2](https://noometry.com/models/gpt-5-2) | [OpenAI](https://noometry.com/providers/openai) | 73.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 18 | [Claude Sonnet 4](https://noometry.com/models/claude-sonnet-4) | [Anthropic](https://noometry.com/providers/anthropic) | 73% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 19 | [GPT-5](https://noometry.com/models/gpt-5) | [OpenAI](https://noometry.com/providers/openai) | 72.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 20 | [Qwen3 Max](https://noometry.com/models/qwen3-max) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 72.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 21 | [o3-pro](https://noometry.com/models/o3-pro) | [OpenAI](https://noometry.com/providers/openai) | 72.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 22 | [Gemini 2.5 Pro](https://noometry.com/models/gemini-2-5-pro) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 70.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 23 | [GPT-5-Codex](https://noometry.com/models/gpt-5-codex) | [OpenAI](https://noometry.com/providers/openai) | 70.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 24 | [GPT-5 Mini](https://noometry.com/models/gpt-5-mini) | [OpenAI](https://noometry.com/providers/openai) | 70.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 25 | [DeepSeek-R1](https://noometry.com/models/deepseek-r1) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 69.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 26 | [Qwen3 235B-A22B](https://noometry.com/models/qwen3-235b-a22b) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 69.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 27 | [o3](https://noometry.com/models/o3) | [OpenAI](https://noometry.com/providers/openai) | 67.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 28 | [o4-mini](https://noometry.com/models/o4-mini) | [OpenAI](https://noometry.com/providers/openai) | 67.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 29 | [Gemini 3.1 Flash Lite](https://noometry.com/models/gemini-3-1-flash-lite) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 67.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 30 | [GPT-5.6 Sol](https://noometry.com/models/gpt-5-6-sol) | [OpenAI](https://noometry.com/providers/openai) | 67% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 31 | [Qwen3-Next 80B-A3B Instruct](https://noometry.com/models/qwen3-next-80b-a3b-instruct) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 66.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 32 | [Grok 4 Fast](https://noometry.com/models/grok-4-fast) | [xAI](https://noometry.com/providers/xai) | 66.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 33 | [Kimi K2 (Jul 2025)](https://noometry.com/models/kimi-k2) | [Moonshot AI](https://noometry.com/providers/moonshot) | 64.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 34 | [GPT-5.4](https://noometry.com/models/gpt-5-4) | [OpenAI](https://noometry.com/providers/openai) | 63.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 35 | [Gemma 4 31B IT](https://noometry.com/models/gemma-4-31b-it) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 63.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 36 | [Qwen Plus](https://noometry.com/models/qwen-plus) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 63.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 37 | [GLM-5.2](https://noometry.com/models/glm-5-2) | [Z.ai (Zhipu)](https://noometry.com/providers/zai) | 62.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 38 | [Step 3](https://noometry.com/models/step-3) |  [![](/logos/stepfun.svg) StepFun](https://noometry.com/providers/stepfun) | 62.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 39 | [GPT-5 Nano](https://noometry.com/models/gpt-5-nano) | [OpenAI](https://noometry.com/providers/openai) | 62.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 40 | [Grok 3](https://noometry.com/models/grok-3) | [xAI](https://noometry.com/providers/xai) | 61.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 41 | [Grok-3 mini](https://noometry.com/models/grok-3-mini) | [xAI](https://noometry.com/providers/xai) | 61.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 42 | [GLM-4.5V](https://noometry.com/models/glm-4-5v) | [Z.ai (Zhipu)](https://noometry.com/providers/zai) | 59.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 43 | [gpt-oss-120b](https://noometry.com/models/gpt-oss-120b) | [OpenAI](https://noometry.com/providers/openai) | 58.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 44 | [Claude Sonnet 4.5](https://noometry.com/models/claude-sonnet-4-5) | [Anthropic](https://noometry.com/providers/anthropic) | 57.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 45 | [GLM-4.5](https://noometry.com/models/glm-4-5) | [Z.ai (Zhipu)](https://noometry.com/providers/zai) | 57.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 46 | [MiniMax-M2](https://noometry.com/models/minimax-m2) |  [![](/logos/minimax.svg) MiniMax](https://noometry.com/providers/minimax) | 57.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 47 | [DeepSeek-V3.1-Terminus](https://noometry.com/models/deepseek-v3-1-terminus) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 57.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 48 | [Gemini 2.5 Flash](https://noometry.com/models/gemini-2-5-flash) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 56.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 49 | [Llama 4 Maverick](https://noometry.com/models/llama-4-maverick) |  [![](/logos/meta.svg) Meta](https://noometry.com/providers/meta) | 55.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 50 | [Hermes 4 405B](https://noometry.com/models/hermes-4-405b) | [Nous Research](https://noometry.com/providers/nous-research) | 55.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 51 | [MiniMax-M2.5](https://noometry.com/models/minimax-m2-5) |  [![](/logos/minimax.svg) MiniMax](https://noometry.com/providers/minimax) | 55.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 52 | [Qwen3-30B-A3B](https://noometry.com/models/qwen3-30b-a3b) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 54.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 53 | [Qwen3 32B](https://noometry.com/models/qwen3-32b) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 54.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 54 | [DeepSeek V4 Pro](https://noometry.com/models/deepseek-v4-pro) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 53.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 55 | [DeepSeek-V3.1](https://noometry.com/models/deepseek-v3-1) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 53.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 56 | [gpt-oss-20b](https://noometry.com/models/gpt-oss-20b) | [OpenAI](https://noometry.com/providers/openai) | 53.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 57 | [DeepSeek-R1-Distill-Llama-70B](https://noometry.com/models/deepseek-r1-distill-llama-70b) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 52.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 58 | [DeepSeek-V3](https://noometry.com/models/deepseek-v3) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 52.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 59 | [GPT-4.1](https://noometry.com/models/gpt-4-1) | [OpenAI](https://noometry.com/providers/openai) | 52.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 60 | [DeepSeek-V3.2-Exp](https://noometry.com/models/deepseek-v3-2-exp) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 52.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 61 | [DeepSeek V4 Flash](https://noometry.com/models/deepseek-v4-flash) |  [![](/logos/deepseek.svg) DeepSeek](https://noometry.com/providers/deepseek) | 52.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 62 | [GPT-5.6 Terra](https://noometry.com/models/gpt-5-6-terra) | [OpenAI](https://noometry.com/providers/openai) | 51.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 63 | [Mistral Large 3](https://noometry.com/models/mistral-large-3) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 50.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 64 | [Mistral Medium](https://noometry.com/models/mistral-medium) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 50% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 65 | [Qwen3-Coder 480B-A35B Instruct](https://noometry.com/models/qwen3-coder-480b-a35b-instruct) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 49.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 66 | [GPT-5.6 Luna](https://noometry.com/models/gpt-5-6-luna) | [OpenAI](https://noometry.com/providers/openai) | 49.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 67 | [Qwen3 14B](https://noometry.com/models/qwen3-14b) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 49.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 68 | [GPT-4.1 mini](https://noometry.com/models/gpt-4-1-mini) | [OpenAI](https://noometry.com/providers/openai) | 48.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 69 | [GLM-4.6](https://noometry.com/models/glm-4-6) | [Z.ai (Zhipu)](https://noometry.com/providers/zai) | 47.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 70 | [Llama 3.1-405B](https://noometry.com/models/llama-3-1-405b) |  [![](/logos/meta.svg) Meta](https://noometry.com/providers/meta) | 45% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 71 | [Nova Premier 1.0](https://noometry.com/models/nova-premier-1-0) | [Amazon](https://noometry.com/providers/amazon) | 44.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 72 | [Longcat Flash Chat](https://noometry.com/models/longcat-flash-chat) | [Meituan](https://noometry.com/providers/meituan) | 43.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 73 | [GLM-4.5-Air](https://noometry.com/models/glm-4-5-air) | [Z.ai (Zhipu)](https://noometry.com/providers/zai) | 43% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 74 | [MiniMax-01](https://noometry.com/models/minimax-01) |  [![](/logos/minimax.svg) MiniMax](https://noometry.com/providers/minimax) | 42.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 75 | [Mistral Medium 3.5](https://noometry.com/models/mistral-medium-3-5) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 41.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 76 | [Gemini 2.5 Flash-Lite](https://noometry.com/models/gemini-2-5-flash-lite) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 40.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 77 | [Mistral Small 4](https://noometry.com/models/mistral-small-4) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 40.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 78 | [Gemma 3 27B](https://noometry.com/models/gemma-3-27b) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 40.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 79 | [Mistral Small 3.2](https://noometry.com/models/mistral-small-3-2) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 40.4% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 80 | [GPT-5.4 nano](https://noometry.com/models/gpt-5-4-nano) | [OpenAI](https://noometry.com/providers/openai) | 39.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 81 | [GPT-5.4 mini](https://noometry.com/models/gpt-5-4-mini) | [OpenAI](https://noometry.com/providers/openai) | 37.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 82 | [Gemini 2.0 Flash (Feb 2025)](https://noometry.com/models/gemini-2-0-flash) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 37.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 83 | [Mistral Small](https://noometry.com/models/mistral-small) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 37.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 84 | [Devstral Medium](https://noometry.com/models/devstral-medium) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 37.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 85 | [Devstral Small 2505](https://noometry.com/models/devstral-small) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 37.7% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 86 | [Llama 4 Scout](https://noometry.com/models/llama-4-scout) |  [![](/logos/meta.svg) Meta](https://noometry.com/providers/meta) | 36.9% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 87 | [Qwen2.5-VL 72B Instruct](https://noometry.com/models/qwen2-5-vl-72b-instruct) |  [![](/logos/alibaba.svg) Alibaba (Qwen)](https://noometry.com/providers/alibaba) | 36% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 88 | [Llama 3-70B](https://noometry.com/models/llama-3-70b) |  [![](/logos/meta.svg) Meta](https://noometry.com/providers/meta) | 35.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 89 | [Claude 3 Haiku](https://noometry.com/models/claude-3-haiku) | [Anthropic](https://noometry.com/providers/anthropic) | 34.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 90 | [GPT-4.1 nano](https://noometry.com/models/gpt-4-1-nano) | [OpenAI](https://noometry.com/providers/openai) | 33.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 91 | [Codestral](https://noometry.com/models/codestral) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 32.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 92 | [Gemma 3n E4b IT](https://noometry.com/models/gemma-3n-e4b-it) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 31.5% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 93 | [Command A](https://noometry.com/models/command-a) |  [![](/logos/cohere.svg) Cohere](https://noometry.com/providers/cohere) | 28.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 94 | [GPT-4o mini](https://noometry.com/models/gpt-4o-mini) | [OpenAI](https://noometry.com/providers/openai) | 28.8% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 95 | [Jamba Large](https://noometry.com/models/jamba-large) | [AI21 Labs](https://noometry.com/providers/ai21) | 26.1% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 96 | [Gemma 3 4B](https://noometry.com/models/gemma-3-4b) |  [![](/logos/google.svg) Google](https://noometry.com/providers/google) | 25.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 97 | [Mercury](https://noometry.com/models/mercury) | [Inception](https://noometry.com/providers/inception) | 21.6% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 98 | [Magistral Medium](https://noometry.com/models/magistral-medium) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 16.2% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| 99 | [Magistral Small](https://noometry.com/models/magistral-small) |  [![](/logos/mistral.svg) Mistral AI](https://noometry.com/providers/mistral) | 6.3% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |

## Compare the leaders

-   [Claude Fable 5 vs Claude Opus 4.8](https://noometry.com/compare/claude-fable-5-vs-claude-opus-4-8)
-   [Claude Fable 5 vs GPT-5.5](https://noometry.com/compare/claude-fable-5-vs-gpt-5-5)
-   [Claude Fable 5 vs Claude Opus 4.6](https://noometry.com/compare/claude-fable-5-vs-claude-opus-4-6)
-   [Claude Fable 5 vs Grok 4.5](https://noometry.com/compare/claude-fable-5-vs-grok-4-5)
-   [Claude Opus 4.8 vs GPT-5.5](https://noometry.com/compare/claude-opus-4-8-vs-gpt-5-5)
-   [Claude Opus 4.8 vs Claude Opus 4.6](https://noometry.com/compare/claude-opus-4-6-vs-claude-opus-4-8)

## Other reasoning benchmarks

-   [ARC-AGI-2](https://noometry.com/benchmarks/arc-agi-2)
-   [SimpleBench](https://noometry.com/benchmarks/simplebench)
-   [NYT Connections (extended)](https://noometry.com/benchmarks/nyt-connections)
-   [ARC-AGI-1](https://noometry.com/benchmarks/arc-agi-1)
-   [CritPt](https://noometry.com/benchmarks/critpt)
-   [Chess Puzzles](https://noometry.com/benchmarks/chess-puzzles)
-   [EnigmaEval](https://noometry.com/benchmarks/enigmaeval)
-   [Thematic Generalization](https://noometry.com/benchmarks/thematic-generalization)
-   [LMArena Hard Prompts](https://noometry.com/benchmarks/arena-hard-prompts)
-   [EBR-Bench](https://noometry.com/benchmarks/ebr-bench)
-   [LiveBench Reasoning](https://noometry.com/benchmarks/livebench-reasoning)
-   [Mystery Game Puzzles](https://noometry.com/benchmarks/mystery-game-puzzles)

## Frequently asked questions

### What does Kagi LLM Benchmark measure?

A private, frequently rotated set of reasoning, coding and instruction-following tasks from Kagi, designed so models cannot be trained on it.

### Which model has the highest Kagi LLM Benchmark score?

As of October 2026, Claude Fable 5 has the highest published Kagi LLM Benchmark score on Noometry at 91.4%, out of 99 models with results.

### What is the best open-weight model on Kagi LLM Benchmark?

Kimi K2.5 has the highest Kagi LLM Benchmark accuracy among open-weight models at 78.5%, ranking 9 of 99 overall.

### Cite this page

Noometry. (2026). Kagi LLM Benchmark leaderboard. Retrieved October 10, 2026, from https://noometry.com/benchmarks/kagi-reasoning

Quote Noometry with a link back to this page. It is also available in [Markdown](https://noometry.com/md/benchmarks/kagi-reasoning.md).