Dashboard

The state of language models

Rankings, releases and benchmark movement · updated October 10, 2026
Models tracked
585
86 organizations
Open weights
381
65% of catalog
Releases · 90 days
48
tracked in the last quarter
Open–closed gap
246.6 Elo
MiMo-V2.6-Pro 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 Opus 5.5Claude Sonnet 5.5GPT-6 AstraMiMo-V2.6-ProMuse Spark 1.3
Latest releases
Model explorer →
Oct 2026Claude Haiku 5.5Anthropic2922.0CLOSEDOct 2026Mistral Large 4Mistral AI2569.3OPENSep 2026Gemini 4 ArgonGoogle3254.0CLOSEDSep 2026GPT-6.1 SolOpenAI3125.4CLOSEDSep 2026Claude Sonnet 5.5Anthropic3240.8CLOSEDSep 2026Claude Opus 5.5Anthropic3266.5CLOSEDSep 2026GPT-6 LunaOpenAI2529.5CLOSEDSep 2026GPT-6 SolOpenAI2848.0CLOSED
Top ranked
All rankings →
1Claude Opus 5.53266.5
2Gemini 4 Argon3254.0
3Claude Sonnet 5.53240.8
4Muse Spark 1.33062.2
5MiMo-V2.6-Pro3019.9
6GPT-6 Astra3012.7
7Claude Fable 5.12997.4
8Qwen3.8-Max-09022988.4
Biggest movers
Elo gain over previous family release
Sarvam-105Bvs Sarvam-1 (2B)+1624Mistral Large 4vs Mistral Large 3+1324.6DeepSeek-V4-Flash-0731vs DeepSeek-V4-Flash (Non-Think)+1038.6MiMo-V2.6-Provs MiMo-V2.5+1015.4SmolLM3 3B (Think)vs SmolLM2 1.7B+1003.7
Quick compare