Anthropic and OpenAI cut prices as new models land hours apart

Anthropic released Claude Opus 5.5 on Sept. 22, and OpenAI followed 90 minutes later with GPT-6 Sol and Luna, cutting prices on both companies' flagship models the same day, according to TechCrunch.

Anthropic cut Opus 5.5's output price to $20 per million tokens, down from $25, and called it "the strongest-performing model we've tested to date," TechCrunch reported. The model beats the larger Fable model on many benchmarks and matches Mythos on biology and cybersecurity tasks, which subjects it to the same safeguards, including rerouting sensitive cybersecurity questions to the smaller Opus 4.8 model, TechCrunch reported. It is Anthropic's first release since CEO Dario Amodei said the company would deliberately pace its capability advances to let safety work keep up, according to TechCrunch.

GPT-6 Sol, built for coding and other complex tasks, and GPT-6 Luna, built for summarization and clerical work, cost 50% less than their GPT-5.6 predecessors, and Sol makes about half as many factual mistakes, OpenAI said, according to TechCrunch. "GPT-6 Astra introduced a new generation of intelligence; these models extend its benefits by making that intelligence more efficient and accessible," OpenAI said, TechCrunch reported. Sol and Luna are rolling out in ChatGPT, Codex and the API, with Luna also going to free and Go-tier users.

The new list prices show how far the cuts go. GPT-6 Luna costs $0.10 per million input tokens and $0.50 per million output tokens, half of GPT-5.6 Luna's price and among the cheapest models OpenAI has ever sold, developer Simon Willison wrote on his blog. GPT-6 Sol runs $2 and $10 per million tokens, matching xAI's Grok 4.7. Anthropic's move to $4 and $20 per million tokens for Opus 5.5 also cut cache read prices 60%, which Willison said matters most for long agent conversations that keep rereading the same context.

For builders picking a model by task and price instead of brand loyalty, the gap between a frontier model and a cheap one keeps shrinking. Willison said the sharpest competition now sits in the middle tier, between Sol, Luna, Grok and Anthropic's own Haiku line, which is overdue for a refresh to stay price competitive.