Sulat.com
AI models
$10 off the fastest Kimi K3 and GLM 5.3
$10 off the fastest Kimi K3 and GLM 5.3 from Synthetic
Vercel AI Gateway logo

Model details

GPT-6 Astra (Fast)

GPT-6 Astra Fast appears in OpenAI's Codex IDE as a bundled model picker item, with its visibility and async-question guidance wired into the Codex CLI behavior. The Codex 0.153.4 hotfix made Astra the bundled default when no model is explicitly configured, and the same release tightened Astra's guidance to use asynchronous clarification only when the relevant tool is available in the session, signaling that the variant is positioned for interactive, tool-driven Codex flows rather than standalone chat use.

Within Codex, Astra Fast is treated as an internal selectable model identifier exposed through the CLI's model catalog rather than as a separately documented foundation model. Subsequent Codex hotfixes back-ported the GPT-6-Astra model catalog into 0.153 and added API-side configuration support without changing the default model, and a later fix corrected the GPT-6-Astra Fast tier description so that Codex reflects the intended speed/usage trade-off for the Fast offering. This makes GPT-6 Astra Fast best understood as the fast Codex IDE flavor of the Astra model family rather than a headless API product.

Vercel AI Gatewayopenai/gpt-6-astra-fastgpt-astra
Open provider pageProvider docs

Quick Info

Powered by
Provider
Vercel AI Gateway
Model key
openai/gpt-6-astra-fast
Release date
Sep 4, 2026
Last updated
Sep 4, 2026
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$20.00
Output token cost
$100.00

Limits

Input tokens
922,000 tokens
Output tokens
128,000 tokens
Context window
1,050,000 tokens
News search byUse it for free

Latest news about GPT-6 Astra (Fast)

Vercel AI Gateway

CoverageRelease Notes

The Releasebot aggregator surfaces the same Codex CLI 0.153.3 and 0.153.4 release notes that explicitly name GPT-6-Astra, corroborating the official changelog's record of the model's integration into OpenAI's developer tooling. The 0.153.3 entry confirms GPT-6-Astra was added to the Amazon Bedrock model picker for Mant Releasebot also reproduces the accompanying async-clarification bug fixes for GPT-6-Astra: 0.153.3 corrected Astra's guidance to use the supported tool and accept text only, while 0.153.4 qualified that asynchronous questions should only be used when the tool is available in the session. While Releasebot is a secondary

Vercel AI Gateway

CoverageRelease Notes

The official ChatGPT & Codex changelog documents three Codex CLI releases (0.153.2 on September 3, 2026 and 0.153.3/0.153.4 on September 4, 2026) that explicitly name GPT-6-Astra. The 0.153.2 entry corrects the GPT-6-Astra Fast tier description, providing the strongest primary-style evidence in the candidate set that t Bug fixes across 0.153.3 and 0.153.4 refine Astra's guidance for asynchronous clarification questions to only fire when the supported tool is available in the session, and clarify that the clarification path accepts text only. These changelog entries establish concrete developer-surface behavior for GPT-6-Astra (includ

Videos about GPT-6 Astra (Fast)

More models around GPT-6 Astra (Fast)