OpenAIOpenAI·text

GPT-6.2 Astra

VisionReasoningWeb searchFunction calling
Advertised as gpt-6.2-astraopenai-gpt-62-astraopenai/gpt-6.2-astra
Quick reference
GPT-6.2 Astra — TLDR
  • 🏢 OpenAI's flagship GPT-6.2 model, released November 17, 2026.
  • 📏 1.05M-token context: 922K input, 128K output tokens.
  • 👁️ Text and image inputs; stronger visual reasoning for sites, slides, 3D.
  • 🔧 Function calling, web search, computer use, document work.
  • 🧠 Built for engineering, professional work, and scientific research.
  • ⚡ OpenAI reports better results using fewer output tokens than before.
  • 🎯 OpenAI reports 62.5% on Artificial Analysis's AA-AnalystAgent task set.
  • 💬 OpenAI says writing is more professional and readable than GPT-6 Astra.
💰 Best price on Antseed
$0.300 / $1.50−96%
per 1M · cheapest in / out
🐜 Sellers
12
advertising on Antseed
🔬 Model checks
Not audited yet
no audit of this model so far
📏 Context
1.1M tokens
Provider

OpenAI is an American artificial intelligence research organization headquartered in San Francisco, structured as both a for-profit public benefit corporation and a nonprofit foundation. The lab developed the GPT family of large language models, the DALL-E image generation…

Explore 37 more models by OpenAI →
About this model

GPT-6.2 Astra is OpenAI's frontier model for the hardest end-to-end projects, launched alongside the efficiency-oriented GPT-6.2 Sol and the higher-compute GPT-6.2 Astra Pro. Per OpenAI's API documentation, it can build and revise software, use a computer, and work with documents, accepting text and images and responding with text, with a 1.05M-token context window (922K input, 128K output) and support for function calling and web search.

Relative to its direct predecessor GPT-6 Astra, OpenAI positions 6.2 as delivering better results on demanding tasks while emitting fewer output tokens, so more work fits inside the same usage limits. OpenAI also says GPT-6.2 Astra's writing is more professional and readable than GPT-6 Astra's, and highlights stronger visual reasoning applied to websites, slideshows, and 3D designs.

On evaluations cited in OpenAI's own launch post, GPT-6.2 Astra answers 62.5% of questions correctly on all five independent attempts in Artificial Analysis's AA-AnalystAgent suite, which works on real spreadsheets and reports, and OpenAI reports a new high for the model on Terminal-Bench-Science 0.1, a benchmark for scientific workflows carried out in a terminal. These figures are self-reported by the provider.

For lineage context: GPT-6 Astra, released in September 2026, was already OpenAI's flagship for complex reasoning, coding, computer use, research, and document creation, with configurable reasoning effort levels. GPT-6.2 Astra continues that line, with long-context requests above 272K input tokens billed at higher rates.

Sources
developers.openai.comGPT-6.2 Astra Model | OpenAI API· developers.openai.comopenai.comIntroducing GPT-6.2 Astra and Sol | OpenAI· openai.com

This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.

Usage on Antseed · GPT-6.2 Astra
Tokens served
142.76M
input + output
Requests
808
settled calls
Buyers
14
distinct, on this model
Sellers used
8
of 12 advertising
Settled
$157.76
gross USDC, this model

Daily usage · GPT-6.2 Astra · Organic sellers

Top sellers · GPT-6.2 Astra · Organic

Sellers serving GPT-6.2 Astra (12)compare on the network explorer →
SellerHealthRep↓SalesTokensRoutingInput $/MCached
$/M
Output $/MSave1st tokentok/s

"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what Antseed Desktop ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.