Vercel adds OpenAI’s GPT-6.2 Astra and Sol to AI Gateway

Vercel has made OpenAI’s GPT-6.2 Astra and GPT-6.2 Sol available through AI Gateway, giving developers access to both models through several APIs, the AI SDK and supported coding agents. Astra targets complex, multi-step work, while Sol is designed for everyday tasks.
- GPT-6.2 Astra and GPT-6.2 Sol are now available on AI Gateway.
- Astra targets complex projects, existing codebases, CAD and simulation tools.
- Sol supports everyday work, agentic workflows, coding and interactive browser experiences.
- Both models work with the AI SDK, OpenAI-compatible APIs and connected coding agents.
- Ultrafast GPT-6.2 Astra output costs 60% less than GPT-6 Astra output.
Two models for different workloads
GPT-6.2 Astra is intended for complex, multi-step projects. According to Vercel, it can produce slides, spreadsheets and documents that follow a user’s templates and the conventions of their field. The model is also better at implementing and testing complex changes in existing codebases.
Its computer-use capabilities are designed for more precise interaction, including work down to individual parts in CAD and simulation tools. That positions Astra for tasks where the model must operate across documents, software and technical environments rather than only generate text.
GPT-6.2 Sol is a general model for everyday work. Vercel says it improves on earlier Sol models for agentic workflows, coding and writing. It can also turn a written brief into an interactive browser experience, including an example such as a playable game.
Access through APIs and coding agents
Both models are available through the AI SDK, OpenAI-compatible Chat Completions and Responses APIs, and coding agents connected to AI Gateway. Developers can select them with the model names openai/gpt-6.2-astra and openai/gpt-6.2-sol.
Vercel’s example uses openai/gpt-6.2-sol with the AI SDK to stream a response for a prompt asking the model to draft a one-page project brief from meeting notes. Developers can also try either model in the model playgrounds for GPT-6.2 Astra and GPT-6.2 Sol.
For coding agents such as Codex and Cursor, Vercel says users should install the latest Vercel CLI and run npx vercel ai-gateway setup. They can then select either model in the agent.
Fast service tiers and gateway controls
OpenAI’s fast mode and ultrafast service tier are available for both models. In the AI SDK, developers can set serviceTier: 'ultrafast' in the OpenAI provider options. Chat Completions and Responses requests use service_tier.
Ultrafast is billed at a higher per-token rate than the default tier. Vercel says ultrafast output on GPT-6.2 Astra costs 60% less than on GPT-6 Astra.
AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover and performance optimizations for higher-than-provider uptime. It also includes custom reporting, Zero Data Retention support, budgets for API keys and routing rules.
What the availability means for developers
The launch gives teams using AI Gateway another pair of OpenAI models without requiring separate integration paths for each supported workflow. Astra is aimed at demanding technical and document-based projects, while Sol provides a broader option for coding, writing and agentic tasks.
Vercel says AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including Bring Your Own Key requests. Developers can also view the full list of language models available through the gateway and use its controls for routing, budgets, usage tracking and failover.
Vercel: GPT-6.2 Astra and Sol now available on AI Gateway

