VerdictPal
Join

Models / OpenAI

GPT-6 Astra

GPT-6 · Released 2026-09-03

OpenAI's 3 Sep 2026 flagship. API list $10 / $50 per 1M tokens, 1.05M context, 128K output, knowledge cutoff 30 Apr 2026. Vals Index 66.61% on the live board; AA Intelligence Index v4.2 is 55 at max effort. ChatGPT ships it as GPT-6 Pro on eligible plans.

Current OpenAI flagship for long-horizon coding, computer use, research, and document work. Prefer Vals and AA ledger rows over vendor FrontierMath / ARC-AGI-3 / ExploitBench headlines.

#8 of 24 on Vals Index · current snapshot
67 Vals Index · 4 benchmark rows · 2 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite60
Context1M
Input / 1M$10
Output / 1M$50
Knowledge cutoff2026-04-30
Output speed64 t/s
TTFT (default API)384.30s
Statusflagship
ChatGPT
Not enough families covered yet.Add more benchmark rows for GPT-6 Astra to see the family profile.
Editor's note
Recorded from the 3 Sep 2026 OpenAI launch post, the API model page, the live AA model page (v4.2 index 55 at max), and the live Vals Index (66.61%). OpenAI's FrontierMath Tier 4, ARC-AGI-3, and ExploitBench figures stay vendor-claimed — not ledger rows. Editorial depth still needs a desk pass.
Benchmark placements

Where GPT-6 Astra places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
AA time to first tokenPerformance#1/ 186Artificial Analysis — TTFT2026-09-05
Vals IndexAgentic#8/ 2367Vals AI — Vals Index2026-09-05
Artificial Analysis Intelligence IndexReasoning#18/ 18755Artificial Analysis — Intelligence Index2026-09-05
AA output speedPerformance#57/ 187Artificial Analysis — output speed2026-09-05
Family context

The three highest-scoring models with pages in each capability family. Where GPT-6 Astra shows up, it's highlighted.

ReasoningGPT-6 Astra · 55
  1. 1
    Claude Mythos Preview
    95
  2. 2
    Gemini 3.7 Flash
    95
  3. 3
    Grok 4.6
    95
AgenticGPT-6 Astra · 67
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Newest receipts

Benchmark rows added to the public ledger for GPT-6 Astra in the last 120 days. Older rows live in the full table below.

4 recent
BenchmarkFamilyScoreSourceDays ago
Vals IndexAgentic66.61%Vals AI — Vals Index1
Artificial Analysis Intelligence IndexReasoning55%Artificial Analysis — Intelligence Index1
AA output speedPerformance64 t/sArtificial Analysis — output speed1
AA time to first tokenPerformance384.30sArtificial Analysis — TTFT1
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning55
Agentic67
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover any dot for name, score, and input price. Keyboard: tab through the top twelve, or use the ranking below.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMetaMiniMaxxAIZhipuXiaomiOtherMeituanCohereNVIDIA

Full benchmark ledger

Every catalog benchmark for GPT-6 Astra. Scores link to the original source; gaps mean no public row exists yet.

4 sourced rows

4 of 37 catalog benchmarks have a sourced row for GPT-6 Astra.

Reasoning

Agentic

Performance

Changelog
  1. GPT-6 Astra GA on the OpenAI API and a phased ChatGPT / AWS rollout. List $10 / $50 per 1M tokens.
Sources