Leaderboard

AI Leaderboard 2026 — Intelligence Index

Composite ranking of 14 consumer AI models, averaged across coding, math, reasoning, and human-preference benchmarks. Higher score = more capable across the board.

11 of 14 models scored · Updated April 2026

Formula: intelligence = (codingNorm + mathNorm + reasoningNorm + arenaNorm) / N
Each benchmark normalized to 0–100 against the strongest score in our dataset. Models scored on at least 3 of 4 benchmarks. Lower bound: 0. Theoretical max: 100.
# Model Score Coding Math Reasoning Arena Standard plan
1
Claude Opus Anthropic · Claude Opus 4.7
98.8 82% 81% 88% 1428 $100/mo Compare →
2
Claude Anthropic · Claude Sonnet 4.6
95.0 78% 78% 84% 1395 $20/mo Compare →
3
ChatGPT OpenAI · GPT-5.2
93.8 72% 84% 79% 1410 $20/mo Compare →
4
Qwen Alibaba · Qwen 3 Max
92.0 72% 85% 74% 1370 Plus Compare →
5
Gemini Google · Gemini 3 Pro
91.6 70% 80% 78% 1402 $20/mo Compare →
6
Microsoft Copilot Microsoft · GPT-5.2 (M365-tuned)
90.6 68% 82% 76% 1382 $20/mo Compare →
7
Grok xAI · Grok 4
88.5 68% 75% 76% 1378 $30/mo Compare →
8
DeepSeek DeepSeek · DeepSeek V3
87.0 70% 72% 73% 1356 API only Compare →
9
Meta AI Meta · Llama 4 Maverick
83.4 64% 70% 70% 1340 Free Compare →
10
Le Chat Mistral · Mistral Large 2
80.4 62% 65% 68% 1320 $14.99/mo Compare →
11
Pi Inflection AI · Pi 3
74.9 55% 62% 1280 Free Compare →
Not benchmarked on standard suites (different product class)
Or skip the choice

One subscription. Every frontier AI model. $14.99/month.

Perspective AI bundles ChatGPT-class, Claude, Gemini, Grok, Copilot and more in one app. Switch mid-conversation. No per-vendor logins, no separate bills.

Launch app →