AI News/AI News

OpenAI ships GPT-6.2 Astra and Sol, cutting Astra API prices by 20%

The two new models target professional work and everyday tasks, with lower costs and mixed independent results that marketing and content teams should weigh.

Key points

  • OpenAI released GPT-6.2 Astra and Sol on the same day to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and in the API.
  • Astra API prices are 20% lower than GPT-6 Astra: $8 per million input tokens, $0.40 cached, $40 output. Sol keeps $2, $0.10 and $10.
  • Per Artificial Analysis, Astra ranks second on its Intelligence Index behind Claude Fable 5.2, but it sits on the cost-performance frontier.
  • Artificial Analysis found coding flat, and Vals AI ranked both models low on legal benchmarks despite OpenAI's legal claims.

Why it matters for marketers: Cheaper Astra and unchanged Sol pricing lower the cost of agent workflows for slides, documents and research, but independent tests show gaps in coding and legal work worth checking first.

OpenAI has released GPT-6.2 Astra, a frontier model for demanding professional work, and GPT-6.2 Sol, a cheaper model for everyday tasks. Both are available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and through the API. Pro 200 subscribers get Ultrafast access for the first time, starting with Sol.

For content teams, OpenAI's pitch centers on deliverables. It says Astra produces slides, spreadsheets and documents that follow supplied templates more closely, down to layout, colors and font sizes, with titles that state the main takeaway. OpenAI also says Astra uses fewer output tokens, and that Sol comes within 0.5 points of Astra on one coding benchmark at about one-fifth the cost.

Independent results are more mixed. Artificial Analysis ranks Astra second on its Intelligence Index, just behind Anthropic's Claude Fable 5.2, while placing OpenAI's models on the cost-performance frontier. It also reports flat coding results and high hallucination rates on very hard factual questions. Vals AI ranked both models well below the top on legal benchmarks, and Epoch AI said Astra lagged Fable 5.2 on real-world research tasks. Online reviews were reportedly sharply divided.

Teams running template-driven decks, reports or research agents should test Sol first, since its pricing is unchanged and it is positioned as the everyday model. Run Astra on a small set of your own tasks before moving workloads, especially anything involving legal review or factual accuracy. OpenAI also says both models carry tighter cybersecurity safeguards, which may restrict some security testing.

Sources

This brief is our own summary. Read the full reporting at the source.

You read about AI automation. We run it.

DocFluence, built by Z Media, writes, designs, films and publishes for your brand every day, in your voice, from one login.