AILet’s talk
Back to AI updates
OpenAINew features

Use GPT-6.1 Sol with Ultrafast mode

This is mainly relevant if your organisation is on a high-tier ChatGPT Enterprise/Edu plan or Pro $500 and uses Codex or ChatGPT Work for heavy coding/automation tasks — it makes those responses noticeably faster, but doesn't add a new capability for everyday users.

Read the original source 2 min read

What changed

Source summary

ChatGPT and Codex now offer an "Ultrafast" mode that speeds up token generation with the GPT-6.1 Sol model in Codex and ChatGPT Work. Ultrafast mode is available on the Pro $500 plan and on eligible Enterprise and Edu plans.

Why it matters to your work

Practical interpretation

This is mainly relevant if your organisation is on a high-tier ChatGPT Enterprise/Edu plan or Pro $500 and uses Codex or ChatGPT Work for heavy coding/automation tasks — it makes those responses noticeably faster, but doesn't add a new capability for everyday users.

When this could help

Worth switching on for production workloads where users are waiting on the response in real time — live coding helpers, interactive agents, multistep workflows — and the speed gain justifies the trade-off versus standard inference. Less relevant for batch, offline, or cost-sensitive workloads.

Ultrafast mode gives GPT-6.1 Sol a faster-inference option aimed at real-time coding, interactive agents, and other latency-sensitive use cases, at a published cost multiplier (6x for credit/PAYG, 8x for included-usage-rate plans) — so the decision is whether the speed gain justifies the price for a given workload.

How to get started

  1. In the ChatGPT model picker, select GPT-6.1 Sol, then choose Ultrafast from the speed dropdown.

    Provider guidance ↗

  2. In the API or Playground, set the speed configuration to Ultrafast under Agent Configuration.

    Provider guidance ↗

  3. Monitor token usage: Ultrafast costs 6x standard on credit/PAYG plans and 8x on included-usage-rate plans.

    Provider guidance ↗

Build it into your routine

SCALA suggested workflow

  1. Latency-sensitive deployments

    Switch to Ultrafast for real-time coding assistants, interactive agents, or user-facing chat where response time directly affects experience, and revert to Standard or Fast for batch or cost-sensitive workloads.

  2. Cost monitoring

    Track Ultrafast token spend against the 6x/8x multiplier to confirm the speed gain is worth the cost for each use case.

What to check

  • Ultrafast costs 6x standard on credit or pay-as-you-go plans, and 8x on included-usage-rate plans — significant for high-volume workloads. Source ↗
  • Currently listed only for GPT-6.1 Sol; availability for other models is not confirmed. Source ↗

Sources & provider guides

Official documentation checked 12 Oct 2026. SCALA prompts and workflows are suggestions, not claims of hands-on testing.

Source & availability

This is a summary of a published provider announcement, with practical context from SCALA’s feed.

Plan, region, workspace and rollout conditions may affect access in Australia. Check the original source for current availability.

learn.chatgpt.com

This update is primarily for technical teams.