Skip to content
Easy Benchmarks
Best ModelsToolsProvidersUpdatesOpen Full Benchmarks
EnglishFrançaisEspañolDeutschPortuguês

Easy Benchmarks

Tell us what you want to achieve. Easy Benchmarks turns current quality, price, and speed data into a simple shortlist.

Benchmark data is attributed to Artificial Analysis. Model metadata and catalog pricing may also use Vercel AI Gateway data when available.

Latest ReportBenchmark GlossaryMethodologyllms.txt

Latest Report

Latest AI Model Benchmark Report

Current leaders, coverage, recent releases, and source notes from the live dataset.

Download CSV
LLM models
575
AI Model Providers
116
Ranking pages
12
Updated
Jul 16, 2026

Current Leader

OverallClaude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) · 59.9CodingGPT-5.6 Sol (xhigh) · 78.3MathGPT-5 (high) · 99.4%Reasoning & KnowledgeGPT-5.6 Sol (max) · 94.1%SpeedMercury 2 · 880.1 tok/sPrice & ValueQwen3.5 4B (Reasoning) · 335

Recent model releases

Inkling (xhigh)Thinking Machines · LLMJul 15, 2026GPT-5.6 Sol (Non-reasoning)OpenAI · LLMJul 9, 2026GPT-5.6 Terra (high)OpenAI · LLMJul 9, 2026GPT-5.6 Luna (Non-reasoning)OpenAI · LLMJul 9, 2026GPT-5.6 Luna (xhigh)OpenAI · LLMJul 9, 2026GPT-5.6 Sol (high)OpenAI · LLMJul 9, 2026GPT-5.6 Terra (Non-reasoning)
Artificial Analysis Math IndexGPT-5.2 (xhigh) · 99
MMLU-ProGemini 3.1 Pro Preview · 91.2%
LiveBench MathematicsGPT-5.6 Sol (max) · 96.2%
Humanity's Last ExamClaude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) · 53.3%
LiveCodeBenchGemini 3 Pro Preview (high) · 91.7%
SciCodeClaude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) · 60.2%
OpenAI · LLM
Jul 9, 2026
GPT-5.6 Terra (xhigh)OpenAI · LLMJul 9, 2026
GPT-5.6 Sol (max)OpenAI · LLMJul 9, 2026
GPT-5.6 Sol (xhigh)OpenAI · LLMJul 9, 2026
GPT-5.6 Luna (low)OpenAI · LLMJul 9, 2026
GPT-5.6 Terra (low)OpenAI · LLMJul 9, 2026