GPT‑5.5 Pro costs $30/$180 tokens
- OpenAI launched GPT‑5.5 Pro in its API and ChatGPT tiers last week, pricing the higher-accuracy model at $30 input and $180 output. - OpenAI lists GPT‑5.5 Pro with a 1,050,000-token context window, 128,000 max output tokens, and no cached-input discount on standard pricing. - The launch extends OpenAI’s pricier “Pro” tier as rivals push coding models harder. (openai.com)
OpenAI’s new GPT‑5.5 Pro costs six times more than base GPT‑5.5 on standard API input and output pricing. (openai.com 1) (openai.com 2) OpenAI says GPT‑5.5 Pro is priced at $30 per 1 million input tokens and $180 per 1 million output tokens. Base GPT‑5.5 is listed at $5 input and $30 output. (openai.com 1) (openai.com 2) On OpenAI’s model pages, GPT‑5.5 Pro is listed with a 1,050,000-token context window and a maximum output of 128,000 tokens. The company also says the Pro model does not offer a cached-input discount. (openai.com 1) (openai.com 2) A token is a chunk of text, so pricing rises with both the amount of material a model reads and the amount it writes back. A larger context window lets developers feed in far longer codebases, documents, or logs in a single call. (openai.com) (openai.com) OpenAI is positioning GPT‑5.5 Pro as the higher-accuracy option rather than the cheaper default. In its launch post, the company said GPT‑5.5 Pro is for “even higher accuracy,” while GPT‑5.5 is the broader flagship for coding and professional work. (openai.com) (openai.com) The price gap also widens OpenAI’s internal ladder. GPT‑5.4 Pro is listed at the same $30 input price but $120 output, while GPT‑5.5 Pro raises output pricing to $180 per 1 million tokens. (openai.com) (openai.com) OpenAI’s launch materials tie the new family to coding and agent-style tasks, where models call tools, browse, and work through multi-step jobs. The company said GPT‑5.5 and GPT‑5.5 Pro were evaluated with internal and external red-teamers before release. (openai.com) (openai.com) For developers, the tradeoff is straightforward: lower per-token efficiency on price, in exchange for a model OpenAI says is tuned for tougher work. The bill now depends less on the sticker price alone and more on whether a harder model finishes the job in fewer attempts. (openai.com) (openai.com)
