OpenAI launches GPT-6.2 Astra and Sol for professional and everyday work
OpenAI’s GPT-6.2 Astra targets complex professional tasks while Sol focuses on efficient everyday and agentic work, with lower token costs and wider ChatGPT and
By Mohammed Sayyaf · · 3 min read
What shipped
OpenAI has introduced GPT-6.2 Astra and GPT-6.2 Sol, positioning one model for demanding professional work and the other for efficient everyday and agentic tasks. The company says both improve results on hard jobs while using fewer output tokens, which matters for teams hitting usage caps and developers watching API spend.
Astra is framed as the quality leader for complex projects: cleaner code, stronger frontends, and documents that follow templates and domain conventions more closely. Sol is the versatile, lower-cost option for day-to-day workflows. For earlier context on leaner reasoning models, see our note on OpenAI leaner reasoning models.
Why it matters for product teams
The story is less about a single leaderboard jump and more about cost-per-completed-task. OpenAI highlights gains on Agents’ Last Exam versus earlier Astra and versus competing frontier models at a fraction of the cost. For startups, that changes whether agent loops are affordable in production — a theme we also track in agentic developer tools funding.
Key product implications:
- Split traffic between a frontier model and an efficient default.
- Cache aggressively — Astra’s pricing emphasizes discounted cached input.
- Measure abandonment and revision rate, not only raw model scores.
How this compares across the market
Teams comparing providers should also read Vocxa’s coverage of Claude Fable 5.2 and Mythos 5.2, Gemini 4 Hydrogen agentic workflows, and Gemini 4 Neon frontier speed. The competitive picture is a portfolio of models, not one winner.
If your product ships into Europe, pair model selection with governance work from our EU AI Act high-risk checklist and earlier EU AI Act compliance checklist coverage.
Availability and pricing signals
According to OpenAI, GPT-6.2 Astra and Sol are available to Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, with API access via gpt-6.2-astra and gpt-6.2-sol. Ultrafast modes expand further into Pro tiers.
Published API list rates (per million tokens) include Astra at $8 input / $0.40 cached input / $40 output, and Sol at $2 / $0.10 / $10. Always verify live pricing on OpenAI’s site before budgeting.
Vocxa take
Multi-model stacks are now the default architecture. Pair a strong planner with a cheaper executor, keep evals tied to your real workflows, and treat token efficiency as a product feature. Browse more launches in our AI News section or search Vocxa for model comparisons.
External source references
Related coverage
OpenAI ships leaner reasoning models for everyday product work
A new wave of smaller reasoning models aims to cut latency and cost while keeping multi-step planning useful for product teams.
By Mohammed Sayyaf · 2 min read
Google expands Gemini workspace agents
Google is expanding Gemini-powered agents across Workspace with deeper Gmail and Docs integrations for research and drafting workflows.
By John Stephen · 2 min read
Anthropic upgrades Claude Fable 5.2 and Mythos 5.2 for hard technical work
Claude Fable 5.2 is broadly available with stronger coding, reasoning, and vision, while Mythos 5.2 stays gated for verified cyber and life-science use cases.
By John Stephen · 2 min read