--- title: GPT-6.1 Sol Ultrafast rolls out in the API, Codex and... description: OpenAI is rolling out an Ultrafast mode for GPT-6.1 Sol across the API, Codex, and ChatGPT Work, described by OpenAI's developer account as near-Astra level... canonical: https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: GPT-6.1 Sol Ultrafast rolls out in the API, Codex and ChatGPT Work | daily.dev og:description: OpenAI is rolling out an Ultrafast mode for GPT-6.1 Sol across the API, Codex, and ChatGPT Work, described by OpenAI's developer account as near-Astra level... og:url: https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8 og:image: https://api.daily.dev/og/posts/HaHaoj4L8.png og:image:alt: GPT-6.1 Sol Ultrafast rolls out in the API, Codex and ChatGPT Work og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # GPT-6.1 Sol Ultrafast rolls out in the API, Codex and ChatGPT Work **[Collections](https://daily.dev/sources/collections)** · 2 min read · 1 upvotes · 0 comments ## Summary OpenAI is rolling out an Ultrafast mode for GPT-6.1 Sol across the API, Codex, and ChatGPT Work, described by OpenAI's developer account as near-Astra level intelligence running up to 8x faster than Sol Standard. Pricing is reported at $12 per million input tokens and $60 per million output tokens. Whether the quality actually holds up on real workloads remains to be tested by users. ## Content OpenAI has started rolling out Ultrafast mode for GPT-6.1 Sol. It is available in the API, Codex, and ChatGPT Work. OpenAI's developer account describes it as "near-Astra intelligence" at up to 8x the speed of Sol Standard. ## Price Ultrafast costs $12 per million input tokens and $60 per million output tokens. @rohanpaul_ai puts the trade-off bluntly: up to 8x faster, 6x more expensive. ## Amazon Bedrock Amazon Bedrock now offers Ultrafast mode for GPT-6.1 Sol too. It targets low-latency uses such as coding assistants, interactive agents, and customer-facing apps. You can reach it through the Bedrock console or supported APIs, with Bedrock's security, access governance, and audit controls. Regions, pricing, and endpoints are listed in the Bedrock documentation. ## Instant steering in Codex The launch came on day 4 of OpenAI's 28-day pledge. @thsottiaux said steering in Codex is now instant. The model reacts faster when you adjust a running task, so you can correct course in real time instead of letting it burn effort in the wrong direction. He says the change pairs well with Ultrafast. That pairing makes sense to me. A fast model that heads the wrong way just gets you to the wrong answer sooner. Being able to redirect it immediately is what makes the speed useful. ## Related OpenAI's Node SDK 7.31.0 also added API support for prewarmed hosted environments. ## Questions this post answers ### What is GPT-6.1 Sol Ultrafast mode and how much faster is it than Sol Standard? Ultrafast mode is a new speed tier for GPT-6.1 Sol that runs up to 8x faster than Sol Standard while targeting intelligence close to the Astra tier. It is rolling out in the API, Codex, and ChatGPT Work. Whether its output quality matches Astra on real workloads has not yet been independently verified. _Developers evaluating faster coding model tiers can track rollout details like this on daily.dev._ ### How much does GPT-6.1 Sol Ultrafast cost per token? GPT-6.1 Sol Ultrafast is priced at $12 per million input tokens and $60 per million output tokens, according to reporting from testingcatalog. This pricing applies to its availability through the API, Codex, and ChatGPT Work. _Teams comparing model pricing before committing budget can follow updates like this on daily.dev._ ## Community take How the wider developer community reacted, aggregated from 2 discussions and 29 comments across x (as of 2026-10-09). **TL;DR:** Reactions focus mostly on instant steering in Codex as the more practical upgrade than raw speed, while the Ultrafast pricing ($12/$60 vs $2/$10 Standard) draws scrutiny over when it's actually worth it. **Sentiment:** 30% positive · 55% mixed · 15% skeptical **The case for** - Instant steering lets people correct a running agent mid-task instead of waiting for a bad run to finish, which several see as the bigger win than speed alone. - Cutting off bad runs early could reduce cost per finished task more than the price premium adds. - 8x speed could be worth the premium for live agent loops and outage debugging where latency matters most. **The pushback** - Ultrafast costs 6x Standard pricing, and some early reports suggested it was gated behind an expensive plan, pushing effective cost far higher. - Skepticism that the advertised 8x speed and $12/$60 figures are just OpenAI's own claims, unverified independently. - Concern about whether speed gains hold past large context windows (100k+) or are just a short-window optimization. - One commenter argues the need to babysit/steer agents mid-run signals a workflow problem rather than a feature worth celebrating. **By community** - x (mixed): Discussion splits between enthusiasm for instant steering as a practical workflow change and skepticism about pricing, verification of claims, and whether speed scales with context. **Hottest debate:** Whether Ultrafast's cost premium is justified by its speed and steering benefits, or just an expensive gimmick gated behind pricing few will use. **Open questions** - Does a mid-task correction stop a write that already started, or only affect the next step? - Does the 8x speed advantage hold past 100k tokens of context, or is it a short-window optimization? - At what point does Ultrafast's cost actually pay for itself versus staying on Standard for agent loops that burn tokens on retries? **Highlights** > @testingcatalog 8x faster matters more than it sounds. instant steering in codex means you correct the agent mid-run instead of waiting on a bad 20-minute task. speed turns review into collaboration > — [i\_Am\_Snow\_Flake on x · 1 points](https://x.com/i_Am_Snow_Flake/status/2108474593749143989) > @testingcatalog The 8x and $12/$60 figures are OpenAI's. I haven't rerun Sol Ultrafast. The check I'd want is whether a mid-task correction stops a write that already started, or only the next step. I haven't tested that on Codex either. > — [cleon\_suo on x](https://x.com/cleon_suo/status/2108471624718758026) > @testingcatalog The Ultrafast math: $12/$60 is 6x Sol Standard ($2/$10) and 1.2x Astra's list price, so you pay Astra money for near-Astra quality at up to 8x Sol's speed. Worth it for live agents and outage debugging, not for batch jobs. > — [xAIProfitHack on x](https://x.com/xAIProfitHack/status/2108479432734515654) > @testingcatalog Instant steering means a bad run gets cut early, which probably moves cost per finished task more than the $12/$60 does. > — [dhillonpost on x](https://x.com/dhillonpost/status/2108480600621695366) > @testingcatalog Task steering is cool in some situations, but we should be moving away from it generally. The need to babysit agents indicates there is something in your workflow that needs tuning. > — [fs\_dillon on x](https://x.com/fs_dillon/status/2108546691892711801) **Source threads** - [x](https://x.com/rohanpaul_ai/status/2108275940782702945) · 0 points · 2 comments - [x](https://x.com/testingcatalog/status/2108465147186470979) · 0 points · 27 comments ## Similar posts on daily.dev - [Amazon Bedrock announces reduced pricing for OpenAI GPT-5.6 Sol](https://daily.dev/posts/amazon-bedrock-announces-reduced-pricing-for-openai-gpt-5-6-sol-wosqgletk) · AWS · 0 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#openai](https://daily.dev/tags/openai), [#openai-codex](https://daily.dev/tags/openai-codex) [View this post on daily.dev](https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"GPT-6.1 Sol Ultrafast rolls out in the API, Codex and ChatGPT Work","url":"https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8"},"datePublished":"2026-10-08T18:36:44.000Z","dateModified":"2026-10-09T13:51:52.156Z","description":"OpenAI is rolling out an Ultrafast mode for GPT-6.1 Sol across the API, Codex, and ChatGPT Work, described by OpenAI's developer account as near-Astra level...","isAccessibleForFree":true,"articleSection":"Collections","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,openai,openai-codex","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"GPT-6.1 Sol Ultrafast rolls out in the API, Codex and ChatGPT Work"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/gpt-6-1-sol-ultrafast-rolls-out-in-the-api-codex-and-chatgpt-work-hahaoj4l8#faq","mainEntity":[{"@type":"Question","name":"What is GPT-6.1 Sol Ultrafast mode and how much faster is it than Sol Standard?","acceptedAnswer":{"@type":"Answer","text":"Ultrafast mode is a new speed tier for GPT-6.1 Sol that runs up to 8x faster than Sol Standard while targeting intelligence close to the Astra tier. It is rolling out in the API, Codex, and ChatGPT Work. Whether its output quality matches Astra on real workloads has not yet been independently verified. Developers evaluating faster coding model tiers can track rollout details like this on daily.dev."}},{"@type":"Question","name":"How much does GPT-6.1 Sol Ultrafast cost per token?","acceptedAnswer":{"@type":"Answer","text":"GPT-6.1 Sol Ultrafast is priced at $12 per million input tokens and $60 per million output tokens, according to reporting from testingcatalog. This pricing applies to its availability through the API, Codex, and ChatGPT Work. Teams comparing model pricing before committing budget can follow updates like this on daily.dev."}}]} ```