OpenAI has expanded its GPT-6 family with two new models, GPT-6 Sol and GPT-6 Luna, focusing on making advanced AI more affordable for developers and businesses.
The new models arrive just hours after Anthropic introduced Claude Opus 5.5, adding another chapter to the increasingly competitive AI model race. OpenAI says the new models combine improved reliability with significantly lower API costs.
What are GPT-6 Sol and Luna?
GPT-6 Sol is designed for more demanding tasks such as coding, reasoning and complex professional work. Luna is aimed at high-volume tasks with clearer objectives, including document summarisation, information extraction and routine questions.
Both models accept text and image inputs through OpenAI's API. They are also being integrated into products such as ChatGPT Work and Codex.
GPT-6 Sol costs $2 per million input tokens
One of the biggest changes is pricing.
OpenAI lists GPT-6 Sol at $2 per million input tokens and $10 per million output tokens for standard usage within the specified context range. GPT-6 Luna is much cheaper at $0.10 per million input tokens and $0.50 per million output tokens.
OpenAI says the new pricing is roughly half that of the previous GPT-5.6 Sol and Luna models.
OpenAI says Sol makes fewer mistakes
OpenAI says GPT-6 Sol makes about half as many mistakes as its predecessor on the company's internal factuality evaluation, based on anonymised real-world conversations where users had flagged errors.
These are OpenAI's own evaluation results, so they should be viewed as company-reported performance rather than an independent benchmark.
The company is positioning Sol as a model that can deliver stronger reasoning without requiring the same operating cost as its larger flagship systems.
The AI model race is getting cheaper
The timing of the launch is notable. Anthropic released Claude Opus 5.5 around 90 minutes before OpenAI announced GPT-6 Sol and Luna.
Both companies are now competing not only on model intelligence but also on price, speed, coding performance and the ability to run AI agents for longer periods.
For developers, lower token prices could make it more practical to use AI for large software projects and automated workflows.
Comments
Post a Comment