Starting September 30 at 8am PT, Arena.ai will make Anthropic’s Claude Sonnet 5.5 available for direct testing. Claude Sonnet 5.5 (High) stays in Direct Mode for 48 hours and remains in Battle and Agent Mode afterward.
Arena.ai says OpenAI’s GPT-6.1 can now be tested in Agent Arena, where scores are still to come, and in Code Arena for WebDev, Text, Vision, Search, and Document. OpenAI says GPT-6.1 Sol offers near-Astra intelligence for a fifth of the price.
Arena.ai compared 34,580 verdicts from 12 models with human votes across 1,460 Arena battles. On average, a model chose its own answer 58% of the time, while people chose that same answer 34% of the time.
Arena.ai has added Claude Sonnet 5.5 by AnthropicAI to Agent Arena and to Battle Mode for WebDev, Text, Vision, and Document. Claude says it is the second Claude 5.5 model, runs more than 30% faster than Sonnet 5, and costs up to 30% less for most work.
OpenAI’s GPT-6 Sol (Medium) can be selected in Arena Direct Mode for the next 24 hours, through Sept. 29 at 9am PT. Arena.ai says it will remain available in Battle and Agent Mode after that window.