Signal

GPT-6.2 Astra costs $8/1M input and $40/1M output tokens, down 20% from GPT-6 Astra; GPT-6.2 Sol keeps GPT-6.1 Sol's $2/1M input and $10/1M output token pricing

First reported by 9to5Mac ·

The signal ●●●○ Compiled by AI from 9to5Mac and Techmeme
Why you might care

GPT-6.2 Astra's price has decreased by 20%, making advanced AI capabilities more accessible for demanding projects.

What happened

OpenAI has released GPT-6.2 Astra and GPT-6.2 Sol, available today for ChatGPT Work and Codex users on Plus, Pro, Enterprise, and Edu plans. GPT-6.2 Astra, recommended for complex tasks, now costs $8 per 1 million input tokens and $40 per 1 million output tokens, a 20% reduction from GPT-6 Astra. GPT-6.2 Sol maintains its previous pricing of $2 per 1 million input tokens and $10 per 1 million output tokens, offering affordability for everyday coding and writing. Both models are designed for increased efficiency, requiring fewer output tokens to complete demanding tasks. Additionally, the $200/month Pro plan gains access to Ultrafast mode with Sol, while the $500/month Pro plan and eligible Enterprise/Edu workspaces can enable Ultrafast mode for both models, which offers up to 4.5x faster token generation at a higher allowance cost.

What it means

The new GPT-6.2 models signal a continued push by OpenAI to balance performance and cost, making their advanced AI more practical for a wider range of users and applications. The significant price reduction for GPT-6.2 Astra, particularly for demanding tasks, suggests a strategy to encourage adoption for complex use cases where efficiency and reduced output token usage can lead to substantial savings. This move could also pressure competitors to optimize their own pricing structures and performance metrics for high-end generative AI.

For developers and businesses integrating AI, the availability of more cost-effective yet powerful models like GPT-6.2 Astra and Sol means enhanced flexibility in managing AI project budgets and expanding the scope of AI-driven solutions. The introduction of Ultrafast mode, even at a higher allowance cost, provides an option for time-sensitive applications, catering to diverse operational needs. Future developments will likely focus on further optimizing token efficiency and exploring tiered pricing models that accommodate varying levels of demand and complexity.

AI-written summary. May contain errors.

AI Crypto