A
AI Benchmarks
Live Data
About Us

Independent AI Benchmarking,
Built for Developers

AI Benchmarks is a free, independent platform that helps developers, researchers, and businesses find the right AI model for their needs — based on real data, not marketing claims.

🎯 Our Mission

The AI landscape moves fast. New models launch every week, pricing changes overnight, and benchmark numbers are scattered across dozens of sources. We built AI Benchmarks to cut through the noise — aggregating independently measured performance data, real-world pricing, and speed metrics into one clean, always-updated dashboard.

What We Offer

📊
Live Benchmark Data
Performance scores updated automatically from Artificial Analysis — the leading independent AI benchmarking service.
💰
Real Pricing
Actual API pricing from OpenRouter, covering input, output, and blended token costs for 100+ models.
🏆
Value Rankings
Our Value Score (Intelligence ÷ Price) helps you find the most cost-efficient model for your use case.
🔍
AI Finder Quiz
Answer a few questions about your needs and budget — we'll recommend the top 3 models for you.
📐
Category Rankings
Separate rankings for Coding, Mathematics, and General Intelligence so you can find the specialist model you need.
⚡
Speed Benchmarks
Output speed in tokens per second and latency data so you can choose the fastest model for real-time applications.

Our Data Sources

Artificial Analysis ↗

Provides independently measured Intelligence Index, Coding, Math, speed and latency benchmarks. All evaluations are run on dedicated hardware with consistent methodology.

OpenRouter ↗

Aggregates API pricing from 100+ model providers including OpenAI, Anthropic, Google, Meta and more. Pricing data is fetched directly from their public API.

🔄 Always Up to Date

Our data pipeline runs automatically every Monday and Thursday via GitHub Actions, pulling the latest benchmark scores and pricing from both data sources. New models are added automatically — no manual curation required.