Easy Benchmarks
Workspace
Overview
Benchmarks
Benchmarks list
Compare
Overall Index
Coding
Math
MMLU-Pro
Speed
Value
LLMs
Audio
Image
Video
Log inSign up

Easy Benchmarks

LLM Performance Index

Compare current LLMs by benchmark, latency, speed, and price.

A compact analysis surface for choosing the right model with evidence, clear tradeoffs, and grounded assistant answers.

Source

Benchmarks come from Artificial Analysis, TIGER-Lab MMLU-Pro, and LiveBench. Technical metadata is enriched from OpenRouter and Vercel AI Gateway when models match.

Artificial AnalysisOpenRouterMMLU-ProLiveBench
Best Overall
59.9
#1
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
Models Tracked
575
54 creators
Snapshot Jul 16, 2026, 12:22 PM
Best Overall
100
domain
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
Best Value
335
880.1 tok/s
Qwen3.5 4B (Reasoning)

Explore Models

Ranking by Artificial Analysis Intelligence Index. Search narrows this leaderboard.

Percentile
566 matching models on Overall

Top 10 by Overall

Artificial Analysis aggregate LLM intelligence score.

Overall

1 metrics

General leaderboard using the Artificial Analysis Intelligence Index.

Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)100

Coding

3 metrics

Coding leaderboard from coding-specific benchmark metrics.

GPT-5.6 Sol (high)99

Math

4 metrics

Math leaderboard from math-specific benchmark metrics.

GPT-5.6 Sol (max)100

Reasoning & Knowledge

4 metrics

Reasoning leaderboard across GPQA, HLE, MMLU-Pro, and related metrics.

Gemini 3.1 Pro Preview100

Speed

2 metrics

Runtime leaderboard using output speed and latency.

NVIDIA Nemotron Nano 12B v2 VL (Reasoning)98

Price & Value

4 metrics

Cost leaderboard using input, output, blended price, and value index.

Qwen3.5 4B (Non-reasoning)99