The company is integrating a new productivity model into its Codex and ChatGPT Work systems to automate complex, long-running workflows across documents and applications. This model, GPT-6.1 Sol, provides performance near that of Astra while costing one-fifth of Astra's standard token prices. OpenAI also reduced the cost of cached inputs by 95% compared to standard pricing.

Availability of the new model depends on the user's specific plan, client, and workspace settings. GPT-6.1 Sol is specifically designed for repetitive tasks involving code, apps, and extensive documentation where high performance is required at a lower operational cost.