Agentic AI Benchmarks

Interactive cost-vs-performance comparisons of frontier AI models. Automatically synced from Artificial Analysis.

Cost vs. Overall Intelligence

Blended price per 1M tokens vs. Overall Intelligence

Most Attractive

High overall intelligence, low cost

Expensive

High overall intelligence, high cost

Budget

Low overall intelligence, low cost

Avoid

Low overall intelligence, high cost

Click any point to open the model page on Artificial Analysis.

Recent Releases

DeepSeek V4 Pro (Max)
DeepSeek · Apr 24, 2026
DeepSeek V4 Flash (Max)
DeepSeek · Apr 24, 2026
DeepSeek V4 Flash (High)
DeepSeek · Apr 24, 2026
GPT-5.5 (xhigh)
OpenAI · Apr 23, 2026
MiMo-V2.5-Pro
Xiaomi · Apr 22, 2026
Kimi K2.6
Kimi · Apr 20, 2026
Claude Opus 4.7 (max)
Anthropic · Apr 16, 2026
Muse Spark
Meta · Apr 8, 2026
GLM-5.1
Z AI · Apr 7, 2026
Grok 4.20 0309 v2
xAI · Apr 7, 2026
Qwen3.6 Plus
Alibaba · Apr 2, 2026
MiniMax-M2.7
MiniMax · Mar 18, 2026
MiMo-V2-Pro
Xiaomi · Mar 18, 2026
GPT-5.4 mini (xhigh)
OpenAI · Mar 17, 2026
NVIDIA Nemotron 3 Super
NVIDIA · Mar 11, 2026
GPT-5.4 (xhigh)
OpenAI · Mar 5, 2026
Gemini 3.1 Pro Preview
Google · Feb 19, 2026
Claude Sonnet 4.6 (max)
Anthropic · Feb 17, 2026
MiniMax-M2.5
MiniMax · Feb 12, 2026
Claude Opus 4.6 (max)
Anthropic · Feb 5, 2026

Data sourced from Artificial Analysis. Last updated: April 25, 2026.

Prices shown are per 1M tokens. Cost to Run is the total expense to evaluate the model on the Artificial Analysis Intelligence Index.