Frontier Log

OpenAI

GPT-6.2 Astra

Released · openai.com · Launch vibes

Tier Astra › Sol › Terra › Luna

Price

Per 1M tokensInputCachedOutput
List priceCache writes $10.00 (30 minutes), $20.00 (12 hours)$8$0.40$40
Over 272K input tokens$16$0.80$60
Price history
DateInputCachedOutput
openai.com · 2 · Corroborated$8$0.40$40
Over 272K input tokens$16$0.80$60

Context window

TokensContext windowInput limitOutput limit
Default1,050,000922,000128,000
Context history
DateContext windowInput limitOutput limit
As of developers.openai.com · 2 · Corroborated1,050,000922,000128,000

Benchmark results

BenchmarkHarness and effortScore
GPQA Diamondno toolsHarness not specifiedEffort not specified97.0%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Terminal-Bench 4.1Stanford University and Laude Institute’s leaderboard, 18 Nov 2026Codexmax effort62.1%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Terminal-Bench Science 0.1Harness not specifiedmax effort77.1%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedlow effort56.3%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedmedium effort72.3%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedhigh effort74.9%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedxhigh effort74.0%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
FrontierCode 2.0Score, all tasks (132)Run by CognitionHarness not specifiedmax effort44.2%Corroborated The same number is printed in 1 other document: 1
Run by CognitionHarness not specifiedlow effort38.1%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Run by CognitionHarness not specifiedmedium effort42.2%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Run by CognitionHarness not specifiedhigh effort43.6%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Run by CognitionHarness not specifiedxhigh effort44.1%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Cognition’s leaderboard, 18 Nov 2026 · in FrontierCode LeaderboardCodexmax effort44.2%Corroborated The same number is printed in 1 other document: 1
FrontierCode 2.0Pass rate, all tasks (132)Cognition’s leaderboard, 18 Nov 2026Codexmax effort48.2%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
SEC-bench Pro 260505pass@1Harness not specifiedmaximum reasoning effort86.75%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Agents' Last Exam v1.1Score, all tasks (122)Setting from Berkeley RDI’s leaderboardHarness not specifiedxhigh effort78.3%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedlow effort65.8%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedmedium effort72.1%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedhigh effort76.0%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Harness not specifiedmax effort77.6%Checked A person checked this number against its source on 20 Nov 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.
Benchmark sources

Reported by OpenAI

Compare

Related

Sources

All models