Skip to content
All systems operational0 AI providers monitored, polled every 2 minutes
Live status
Intelligence Index

TensorFeed Intelligence Index, October 2026

In progress

As of October 19, 2026, with the month in progress, Claude Opus 5.5 (Anthropic) leads the TensorFeed Intelligence Index for October 2026 with a score of 97.1, among the 17 models ranked on the 2026 benchmark set. Claude Sonnet 5.5 (93.7) and Claude Fable 5.1 (89.9) follow.

Updated daily; latest run . Methodology 1.4. 17 models ranked on the 2026 benchmark set, 29 on the 2024 set, ranked separately. The record freezes at the last daily run of October 2026.

Ranking, 2026 benchmark set

Terminal-Bench 4.0, FrontierCode, OSWorld 2.0, BrowseComp, and Humanity's Last Exam with tools.

RankModelProviderTFII
1Claude Opus 5.5Anthropic97.1
2Claude Sonnet 5.5Anthropic93.7
3Claude Fable 5.1Anthropic89.9
4Claude Opus 5Anthropic83.9
5GPT-6 AstraOpenAI83.3
6Claude Fable 5Anthropic81.6
7GPT-5.6 SolOpenAI74.5
8GLM-5.3Z.ai71.8
9Claude Opus 4.8Anthropic67.9
10Kimi K3Moonshot AI65.1
11GLM-5.3 FlashZ.ai62.8
12GPT-5.6 TerraOpenAI60.9
13Claude Sonnet 5Anthropic59.2
14GPT-5.6 LunaOpenAI53.2
15Gemini 3.7 FlashGoogle52.3
16InklingThinking Machines35.7
17MiniMax M3MiniMax29.8
Ranked separately: 29 models scored only on the 2024 benchmark set

MMLU-Pro, HumanEval, GPQA Diamond, MATH, and SWE-bench Verified. These scores are not comparable with the 2026 set above.

  1. 1. Grok 4.6 (xAI)83.2
  2. 2. Grok 4.5 (xAI)80.0
  3. 3. Gemini 3.8 Flash (Google)79.0
  4. 4. Gemini 3.6 Flash (Google)78.1
  5. 5. Muse Spark 1.1 (Meta)78.1
  6. 6. Gemini 3.5 Flash (Google)77.8
  7. 7. Qwen3.8-Max (Alibaba)77.4
  8. 8. GLM-5.2 (Z.ai)75.6
  9. 9. Qwen3.8 27B (Alibaba)75.2
  10. 10. Gemini 3.5 Flash-Lite (Google)73.3
  11. 11. GPT-5.5 (OpenAI)69.8
  12. 12. Claude Opus 4.7 (Anthropic)69.8
  13. 13. DeepSeek V4 Pro (DeepSeek)69.1
  14. 14. Claude Opus 4.6 (Anthropic)67.8
  15. 15. Claude Sonnet 4.6 (Anthropic)64.6
  16. 16. Gemini 2.5 Pro (Google)64.3
  17. 17. o1 (OpenAI)62.8
  18. 18. DeepSeek V4 Flash (DeepSeek)62.1
  19. 19. Llama 4 Scout (Meta)61.8
  20. 20. Gemini 2.0 Flash (Google)60.7
  21. 21. GPT-4.5 (OpenAI)59.5
  22. 22. o3-mini (OpenAI)58.6
  23. 23. Claude Haiku 4.5 (Anthropic)58.5
  24. 24. DeepSeek V3 (DeepSeek)58.4
  25. 25. Mistral Large (Mistral)57.2
  26. 26. Nemotron 3.5 Lightning (NVIDIA)56.7
  27. 27. Llama 4 Maverick (Meta)56.1
  28. 28. GPT-4o (OpenAI)55.6
  29. 29. Mistral Small (Mistral)55.3

Cite this record

TensorFeed Intelligence Index, October 2026 (in progress, as of October 19, 2026; methodology 1.4). https://tensorfeed.ai/intelligence/2026-10

Machine-readable: https://tensorfeed.ai/api/intelligence/monthly/2026-10. A model's live rank is also available as an embeddable badge.

How the index is built, and why the 2024 and 2026 sets are never compared: TensorFeed Intelligence Index methodology.