OpenAI has released its latest flagship model GPT-5.5, with a focus on stronger agentic capabilities, coding accuracy, and autonomous computer use. The model is now available in ChatGPT and Codex for ChatGPT Plus, Pro, Business, and Enterprise users, while GPT-5.5 Pro is rolling out to higher-tier plans. In benchmarks, GPT-5.5 scored 84.9% on GDPval for knowledge work across 44 occupations, 78.7% on OSWorld-Verified for autonomous computer use, and 98.0% on Tau2-bench Telecom for complex customer service workflows without prompt tuning. OpenAI says it also outperforms GPT-5.4, Anthropic Opus 4.7, and Google Gemini 3.1 Pro on programming tasks involving planning, tool use, and iterative work. In Codex, GPT-5.5 is designed to deliver more accurate single-pass results, preserve context better, check assumptions, and handle ambiguous debugging issues. Developers can access it through the Responses API and Chat Completions API, with a 1 million token context window, pricing at $5 per 1 mill...
Related
Coinbase for Agents can manage trading, payments, and execute workflows on your behalf
Coinbase has introduced Coinbase for Agents, enabling users to connect AI agents directly to their Coinbase accounts for trading, payments, and other financial tasks. These agents ...
Audirvāna Origin 3.0 launches with new Allegro design and UI updates
Audirvāna Origin 3.0 is now available, introducing the Allegro design for greater consistency across computers, tablets, and smartphones. The update adds clearer navigation, contex...
Brave for Android adds custom search engine support in v1.91 update
Brave for Android users can now add custom search engines or searchable sites directly from the app’s settings, enhancing search flexibility. Added engines are accessible as Quick ...