OpenAI unveils GPT-5.5, claims a "new class of intelligence" at double the API price

OpenAI launched GPT-5.5 and GPT-5.5 Pro for paying users and APIs, positioning them as agentic models that can plan, use tools, and handle multi-step workflows.

The models show strong results on coding, math, and long-context benchmarks, with a 1M-token context window and support across ChatGPT, Responses API, Chat Completions API, and Codex.

Independent analysis says the effective cost is higher than GPT-5.4, and hallucinations remain a key weakness.

The release underscores OpenAI’s push toward more autonomous work models, while cost and reliability concerns keep competition from Anthropic Claude and Google Gemini close.