OpenAI has released GPT-6.1 Sol, an upgraded mid-tier model within its GPT-6 family designed for agentic coding, computer interaction, and workflow automation. The company claims GPT-6.1 Sol achieves performance close to its flagship GPT-6 Astra model while pricing input and output tokens at one-fifth of Astra’s standard rates.

The model is available immediately via the OpenAI API under the name gpt-6.1-sol, as well as in ChatGPT Work and Codex. API pricing is established at $2 per million input tokens, $10 per million output tokens, and $0.10 per million cached input tokens. The 95% discount on cached tokens is specifically targeted at multi-step AI agents that repeatedly resend system prompts and context histories.

OpenAI also announced plans to release a GPT-6.1 Sol Ultrafast variant within Codex, promising up to eight times faster token generation speeds. The release heightens competition against mid-tier offerings from Anthropic and Google across developer tools and enterprise API workloads.

Why it matters

  • Deep discounts on cached tokens directly lower operational overhead for multi-step autonomous AI agents.

  • Near-flagship performance at mid-tier prices lowers the cost threshold for enterprise agentic software deployment.

  • High-speed inference variants like Sol Ultrafast target developer workflows where low latency is critical.

Source: marktechpost.com