OpenAI Launches GPT-6.2 Astra and Sol Models for Paid Users
OpenAI has released GPT-6.2 Astra and GPT-6.2 Sol, new AI models available to paid ChatGPT subscribers and API users. Astra targets complex professional tasks with superior performance and lower token costs, while Sol focuses on everyday efficiency. Both models are not available to free-tier users and represent a shift in OpenAI's access strategy.
Key points
- OpenAI, the AI research company, has released two new models, GPT-6.2 Astra and GPT-6.2 Sol, available to ChatGPT Plus, Pro, Enterprise, and Edu users.
- GPT-6.2 Astra is designed for complex professional work, scoring 5.2 points higher than GPT-6 Astra on the Agents' Last Exam at less than a third of the cost.
- GPT-6.2 Sol is positioned as an efficient model for everyday tasks, offering better results with fewer output tokens to maximize usage limits.
- Both models are accessible via the OpenAI API and ChatGPT Work mode but remain unavailable to users on the free tier.
- OpenAI claims Astra sets new benchmarks in scientific work, outperforming previous models on Terminal-Bench-Science 0.1 and the Vals Science Index.
New Frontier Models Released
OpenAI has officially introduced two new additions to its GPT-6.2 family: GPT-6.2 Astra and GPT-6.2 Sol. These models are now available for use by subscribers to ChatGPT’s paid tiers, including Plus, Pro, Enterprise, and Edu plans. The release marks a significant step in OpenAI’s strategy to differentiate its offerings based on user complexity and cost efficiency, targeting both high-end professional workflows and daily productivity tasks.
Astra Targets Complex Professional Work
GPT-6.2 Astra is positioned as the frontier model for the most demanding projects. OpenAI states that it delivers superior results on complex tasks while using fewer output tokens, allowing users to accomplish more within their usage limits. The model is designed to produce slides, spreadsheets, and documents that adhere more closely to provided templates and field-specific conventions, reducing the characteristic "AI" feel of previous iterations.
In terms of performance, OpenAI reports that GPT-6.2 Astra sets a new high on the Agents’ Last Exam, a benchmark evaluating long-horizon professional workflows. The model scored 5.2 points higher than its predecessor, GPT-6 Astra, at less than one-third of the estimated API cost per task. Additionally, it outperformed Anthropic’s Claude Fable 5.2 by 5.5 points at just 5% of the cost. OpenAI also highlights Astra’s capabilities in scientific research, noting it sets new highs on Terminal-Bench-Science 0.1 and the Vals Science Index, which tests the interpretation of real biological data.
Sol for Everyday Efficiency
Complementing Astra is GPT-6.2 Sol, a model designed for versatility and everyday work. While Astra focuses on depth and complexity, Sol is optimized for efficiency, delivering better results on demanding tasks with fewer output tokens. This optimization is intended to help users maximize their usage limits, making it a cost-effective option for routine professional and personal tasks.
Accessibility and Availability
Both GPT-6.2 Astra and GPT-6.2 Sol are available starting today. They can be accessed through ChatGPT Work mode and the OpenAI API. However, access is restricted to paid subscribers. Users on the free tier of ChatGPT will not have access to these new models. Furthermore, the models are only available in Work mode, not in the standard Chat mode, indicating a specific focus on professional and developer use cases rather than casual conversation.
Practical Applications and Developer Showcase
To demonstrate the capabilities of GPT-6.2 Astra, OpenAI developer Thomas Ricouard created a project called Paris Atlas. The model combined street maps with French survey data on more than 131,000 buildings to recreate Paris at full scale. This allowed users to walk or drive through the city in a browser, using screenshots to instruct the model on desired changes. This example highlights the model’s potential in complex spatial and data integration tasks.
Market Implications
The release of GPT-6.2 Astra and Sol underscores the intensifying competition in the AI model market. By offering models that claim superior performance at lower costs compared to competitors like Claude Fable 5.2, OpenAI aims to retain and attract enterprise and developer users. The distinction between Astra and Sol allows OpenAI to cater to different segments of its paid user base, potentially increasing engagement and usage among professionals who require specialized tools.
What Happens Next
As these models roll out to paid users, the focus will shift to how they perform in real-world professional environments. The emphasis on token efficiency and cost reduction suggests that OpenAI is responding to user feedback regarding usage limits and expenses. The availability of these models via the API also indicates that developers will be crucial in integrating GPT-6.2 capabilities into broader applications, potentially driving further adoption in enterprise settings.
Key takeaways
- OpenAI has released GPT-6.2 Astra and Sol for paid ChatGPT and API users.
- Astra is optimized for complex professional tasks and scientific work with high benchmark scores.
- Sol is designed for everyday efficiency, offering better results with fewer tokens.
- Both models are unavailable to free-tier users and restricted to Work mode.
- The release highlights OpenAI's focus on cost efficiency and professional-grade performance.
Sources
The WireByte editorial team synthesises technology news from multiple primary sources, verifies the facts, and links every source. Articles are produced with AI assistance and reviewed under our editorial policy.