Skip to content

Intelligence Index 2026

Objective benchmark comparisons for the world's leading LLMs. Updated weekly as new models and variants are released.

Data last verified: September 6, 2026 · 643 model variants tracked

License Type
Showing 643 of 643 models
AI Model Intelligence Index 2026 — benchmark comparison of 643 LLM variants across Code Arena, Chat Arena, GPQA Diamond, SWE-Bench, and ARC-AGI-2 metrics
#
Model
Price
01
56.8%81.6%$20.00 / 1M tokens
02
54.7%76.9%$20.00 / 1M tokens
03
54.3%75.9%$20.00 / 1M tokens
04
54.1%78%$10.00 / 1M tokens
05
53.8%80.7%$20.00 / 1M tokens
06
53.4%77%$10.00 / 1M tokens
07
53.4%77.1%$20.00 / 1M tokens
08
53.2%76.5%$20.00 / 1M tokens
09
meta logo
Muse Spark 1.3 (max)Meta
53%76.3%$2.00 / 1M tokens
10
52.2%76.7%$20.00 / 1M tokens
11
52%76.5%$10.00 / 1M tokens
12
51.7%79.1%$20.00 / 1M tokens
13
meta logo
Muse Spark 1.3 (xhigh)Meta
51.6%76.5%$2.00 / 1M tokens
14
51.3%77.4%$8.00 / 1M tokens
15
50.6%76.8%$3.00 / 1M tokens
16
50.2%76.2%$6.00 / 1M tokens
17
49.9%77.1%$20.00 / 1M tokens
18
49.8%78.3%$8.00 / 1M tokens
19
49.5%74.3%$10.00 / 1M tokens
20
49.3%75.7%$20.00 / 1M tokens
Page 1 of 33 | 643 models total
...

Complete AI Comparison Library

Deep-dive into technical performance metrics and head-to-head architectural analysis for all leading AI models.

94Comparisons Available