Dashboard

The state of language models

Rankings, releases and benchmark movement · updated July 2026
Models tracked
544
85 organizations
Open weights
367
67% of catalog
Releases · 90 days
54
tracked in the last quarter
Open–closed gap
224.2 Elo
Kimi K3 leads open
Quality vs. price
Frontier Elo against output price, log scale
Pareto frontier
050010001500200025003000$0.1$1$10$100Pareto frontier — best Frontier Elo at each priceClaude Fable 5Claude Opus 5GPT-5.4 ProGPT-5.6 SolKimi K3
Latest releases
Model explorer →
Jul 2026Claude Opus 5Anthropic3290.9CLOSEDJul 2026Gemini 3.5 Flash-LiteGoogle2137.7CLOSEDJul 2026Gemini 3.6 FlashGoogle2618.8CLOSEDJul 2026Laguna S 2.1Poolside2605.7OPENJul 2026Kimi K3Moonshot AI3066.7OPENJul 2026InklingThinking Machines Lab2279.4OPENJul 2026Soofi SSOOFI Consortium763.4OPENJul 2026GPT-5.6 LunaOpenAI2677.2CLOSED
Top ranked
All rankings →
1Claude Opus 53290.9
2Kimi K33066.7
3GPT-5.6 Sol2954.3
4Claude Fable 52950.5
5GPT-5.4 Pro2794.0
6Claude Opus 4.82789.5
7Claude Opus 4.72724.0
8GPT-5.6 Terra2720.8
Biggest movers
Elo gain over previous family release
Sarvam-105Bvs Sarvam-1 (2B)+1630.8SmolLM3 3B (Think)vs SmolLM2 1.7B+1005Hy3vs Hunyuan-A13B+945.3SmolLM3 3B (No Thinking)vs SmolLM2 1.7B+812.1Phi-4-reasoningvs Phi-4-mini (3.8B)+757.5
Quick compare