# OpenAI Puts GPT-6.1 Sol on Ultrafast at $12/$60 per Million Tokens, Selling Up to 8x Speed Across the API, Codex and ChatGPT Work

Times of AI Desk · 2026-10-09 · Models

[https://timesof.ai/2026/10/openai-ultrafast-gpt-6-1-sol](https://timesof.ai/2026/10/openai-ultrafast-gpt-6-1-sol)

> OpenAI said on Oct. 8 that Ultrafast, its fastest service tier, is rolling out for GPT-6.1 Sol in the API, Codex and ChatGPT Work, at up to 8x the speed of Sol Standard. API pricing is $12 per million input tokens and $60 per million output tokens, which OpenAI calls 1.2x the cost of Astra. The API tier is open to all API users with separate rate limits; in Codex and ChatGPT Work it is limited to Pro 500, eligible usage-based Enterprise and credit-based Edu plans. The speed figure is OpenAI's own.

OpenAI is now selling speed on Sol as a priced tier rather than a preview. On **October 8**, it said **Ultrafast** is rolling out for **GPT-6.1 Sol** in the **API**, **Codex** and **ChatGPT Work**, running at up to **8x** the speed of Sol Standard. In August, Ultrafast was a [limited customer preview on GPT-5.6 Sol](https://timesof.ai/2026/08/openai-ultrafast-gpt-5-6-sol-cerebras); the API guide now says Ultrafast for GPT-6 Astra and GPT-6.1 Sol "is available to all API users."

## What OpenAI announced

| | GPT-6.1 Sol, Ultrafast |
| --- | --- |
| Speed | Up to 8x faster than Sol Standard (OpenAI) |
| API price | $12 per million input tokens, $60 per million output tokens |
| OpenAI's comparison | "just 1.2x the cost of Astra" |
| API access | All API users; separate rate limits from Standard and Fast |
| Codex and ChatGPT Work | Pro 500, eligible usage-based Enterprise, credit-based Edu; Enterprise admins must enable it |
| Data residency | US and EU, plus global processing |

OpenAI pitches the tier at "debugging an outage, agents navigating apps, and live experiences where every second counts," and describes it as "near-Astra intelligence." Its API guide recommends WebSockets for agentic workloads, since network overhead on separate requests can eat into the latency gains.

## Rate limits

Default Ultrafast token limits, per OpenAI's API guide:

| API usage tier | GPT-6.1 Sol (TPM) | GPT-6 Astra (TPM) |
| --- | --- | --- |
| Build | 1,000,000 | 500,000 |
| Launch | 4,000,000 | 1,000,000 |
| Grow | 40,000,000 | 5,000,000 |

GPT-6 Astra Ultrafast supports US data residency and global processing only. With the same post, OpenAI added EU data residency for GPT-6.1 Sol Fast and GPT-6 Luna Fast.

## Claims and checks

The up-to-8x figure and "near-Astra intelligence" are OpenAI's descriptions; OpenAI gives no tokens-per-second figure or benchmark for this tier in the announcement, and we have no independent speed measurement. Ultrafast is a service tier for an existing model, not a new checkpoint, so it does not change any model's leaderboard position. The "1.2x the cost of Astra" comparison is OpenAI's wording; the announcement does not say which Astra pricing tier it compares against.

## Limits

- Details come from OpenAI's post in the Announcements section of its developer forum and from its API guide; pricing for cached input and cache writes, which the guide links to a separate pricing table, is not covered here.
- Speed and quality claims are OpenAI's and unverified.

## Sources

- [OpenAI Developer Community: Ultrafast is rolling out today for GPT-6.1 Sol in the API, Codex, and ChatGPT Work (October 8, 2026)](https://community.openai.com/t/ultrafast-is-rolling-out-today-for-gpt-6-1-sol-in-the-api-codex-and-chatgpt-work/1404475/1)
- [OpenAI API docs: Ultrafast mode (accessed October 12, 2026)](https://developers.openai.com/api/docs/guides/ultrafast-mode)
- [Times of AI: Ultrafast Puts Sol at 750 tok/s — a Speed Class, Not a New Checkpoint](https://timesof.ai/2026/08/openai-ultrafast-gpt-5-6-sol-cerebras)