Skip to content
All systems operational0 AI providers monitored, polled every 2 minutes
Live status

GPT-6.2 Astra

Flagship

by OpenAI

As of November 18, 2026, GPT-6.2 Astra by OpenAI costs $8.00 per 1M input tokens and $40.00 per 1M output tokens, with a 1.05M-token context window.

GPT-6.2 Astra is the OpenAI flagship as of November 17, 2026, shipping as gpt-6.2-astra alongside GPT-6.2 Sol and arriving nearly two months after OpenAI scrapped GPT-6.1 Astra for straying outside its authorized scope. The rate card moves down for the first time on an Astra: $8 per million input tokens and $40 output, 20 percent under GPT-6 Astra at $10 and $50, with cached input at $0.40 rather than $1, a $10 cache write, and Batch at half the standard rate. The long-context cliff that defines this generation stays in place: a prompt above 272,000 input tokens bills the whole request at $16 input, $0.80 cached, and $60 output, so a browsing agent that accumulates screenshots crosses into the higher tier without deciding to. Ultrafast bills at 3 times the standard rate through February 28, 2027 and 6 times from March 1, 2027, and the Responses API beta added two hour and eight hour background completion windows priced 30 and 40 percent below standard. The shape is the familiar one: a 1,050,000 token context window with a documented maximum of 922,000 input tokens, 128K output, text and image input, text output only, a May 18, 2026 knowledge cutoff, and reasoning tokens with an effort setting of none rejected. Two API changes can break ported code: tool calling now requires the Responses API, and the Responses API changed how it preserves reasoning when a conversation prefix is edited and how it handles the top-level instructions field. OpenAI publishes its gains as cost ratios rather than a conventional table: Agents' Last Exam 5.2 points above GPT-6 Astra at under a third of the cost per task and 5.5 points above Claude Fable 5.2 at about 5 percent of its cost, Harvey LAB-AA at more than double Fable 5.2's pass rate for about a sixth of the cost, GDPval v2 up 6.2 points while spending 22 percent fewer output tokens, Terminal-Bench-Science 9 points above GPT-6 Astra, and 7 of 68 solved on the Erdos set of FrontierMath against a previous record of 3 of 68. ExploitBench clears 50 percent with safeguards off. Independent measurement is more mixed, and the coding row is the one to read first: Artificial Analysis places Astra second on its Intelligence Index at 62, behind Claude Fable 5.2 at 65 and above Gemini 4 Carbon at 60 and Claude Opus 5.5 at 59, but reports its Coding Agent Index flat against GPT-6 Astra, so coding capability did not move with the price. It leads AA-3D on human-judged Elo at 1318 against 1271 for Fable 5.2, posts 79 on AA-Robotics v0.1, and hallucinates on 59 percent of AA-Omniscience questions at max effort. Epoch's Capabilities Index puts it at 174, tied with Fable 5.2, and Vals ranks it seventh at $10.68 per task. In ChatGPT it reaches Plus, Pro, Enterprise, and Edu users through ChatGPT Work and Codex. GPT-6 Astra is not deprecated and stays on the price list at $10 and $50.

Input Price

$8.00

per 1M tokens

Output Price

$40.00

per 1M tokens

Context Window

1.1M

tokens

Released

2026-11

API access

Capabilities

textvisiontool-usecodereasoning

Key Strengths

  • ✓$8/$40 per 1M tokens, 20 percent under GPT-6 Astra
  • ✓Cached input at $0.40, down 60 percent from $1 on GPT-6 Astra
  • ✓Second on the Artificial Analysis Intelligence Index at 62, behind Claude Fable 5.2 at 65
  • ✓First on AA-3D human-judged Elo at 1318, against 1271 for Fable 5.2
  • ✓1,050,000 token context window with 128K output
  • ✓Background completion windows at 30 and 40 percent below standard token rates

Best For

  • ▸Long-horizon agentic work where cost per task decides the model
  • ▸Computer use and browser-driving agents
  • ▸Legal and professional knowledge work on the Harvey LAB-AA pattern
  • ▸Frontier mathematics and scientific reasoning

Pricing Details

Input tokens

$8.00

per 1M tokens

Output tokens

$40.00

per 1M tokens

Estimated cost per 1K requests

$28.00

~1K input + ~500 output tokens avg

Prices are subject to change. Check the official documentation for current pricing. See the cost calculator for detailed estimates.

Quick answers

How much does GPT-6.2 Astra cost?
As of November 18, 2026, GPT-6.2 Astra costs $8.00 per 1M input tokens and $40.00 per 1M output tokens on the OpenAI API.
What is the context window of GPT-6.2 Astra?
GPT-6.2 Astra has a 1.05M-token context window (1,050,000 tokens).

Related Models

View DocumentationCompare ModelsCost CalculatorFull Pricing Guide