OpenAI launches GPT-6 Sol and Luna, boasting lower cost and fewer mistakes
OpenAI says GPT-6 Sol and Luna bring Astra-level reliability into cheaper, smaller models.
The new Sol and Luna versions cost half as much through the API as the prior 5.6 series, which OpenAI credits to caching and inference gains. Sol is aimed at complex work such as coding, while Luna is positioned for high-volume clerical tasks like summaries, extraction, and quick answers. OpenAI says Sol makes about half as many factual mistakes as its predecessor on its internal evaluation and has a lower coding error rate. The models are rolling out across paid ChatGPT Work, Codex, the API, and, for Luna, Free and Go users.
TechCrunch AI's note
The new Sol and Luna versions cost half as much through the API as the prior 5.6 series, which OpenAI credits to caching and inference gains. Sol is aimed at complex work such as coding, while Luna is positioned for high-volume clerical tasks like summaries, extraction, and quick answers. OpenAI says Sol makes about half as many factual mistakes as its predecessor on its internal evaluation and has a lower coding error rate. The models are rolling out across paid ChatGPT Work, Codex, the API, and, for Luna, Free and Go users.
TechCrunch AI's note
score 9