🇺🇸 English 🇰🇷 한국어 🇯🇵 日本語
Home › LLM & Chatbots
LLM & Chatbots

GPT-6.2 Astra Cuts API Price 20%, Adds Sol

OpenAI released GPT-6.2 Astra and GPT-6.2 Sol on Nov. 17, 2026, cutting Astra's API price 20% to $8 input and $40 output per million tokens while Sol keeps $2/$10. OpenAI says Sol came within half a point of Astra on FrontierCode 2.0 at about one-fifth the cost; benchmarks are vendor-run.

OpenAI on Tuesday, Nov. 17, 2026, released GPT-6.2 Astra and GPT-6.2 Sol in ChatGPT Work, Codex and its API, and cut Astra's list price by 20% to $8 per million input tokens and $40 per million output tokens. Sol, the cheaper everyday model, keeps GPT-6.1 Sol's $2 and $10 rates.

Multistory office building at 1515 Third Street in San Francisco's Mission Bay neighborhood
1515 Third Street in San Francisco's Mission Bay, which housed OpenAI's headquarters when this file photo was taken in June 2025. 사진: Coolcaesar / 위키미디어 공용(CC BY 4.0)

Key takeaways

  • GPT-6.2 Astra lists at $8 input, $0.40 cached input and $40 output per million tokens, down from $10, $1.00 and $50 for GPT-6 Astra.
  • OpenAI says Sol came within half a point of Astra on the FrontierCode 2.0 coding test at about one-fifth of the cost.
  • Both models reached GitHub Copilot the same day, and OpenAI says it launched them with tighter safeguards for higher-risk cybersecurity requests.

GPT-6.2 Astra and Sol API pricing

OpenAI's API pricing page lists the new models as gpt-6.2-astra and gpt-6.2-sol. The biggest change is cached input for Astra, which falls 60%, from $1.00 to $0.40 per million tokens. That matters for agents that resend long instructions and project files on every turn.

OpenAI API standard rates, USD per 1 million tokens (as of Nov. 18, 2026)
ModelInputCached inputOutput
GPT-6.2 Astra$8.00$0.40$40.00
GPT-6 Astra$10.00$1.00$50.00
GPT-6.2 Sol$2.00$0.10$10.00
GPT-6.1 Sol$2.00$0.10$10.00

The page adds that eligible background requests on both GPT-6.2 models are billed 30% below standard rates for a two-hour completion window and 40% below for an eight-hour window.

GPT-6.2 benchmarks: what OpenAI claims

All performance figures so far come from OpenAI. As reported by Digital Trends, the company says GPT-6.2 Astra scored 5.2 percentage points higher than GPT-6 Astra on Agents' Last Exam while costing less than a third as much per task. It also gained nine points on Terminal-Bench Science 0.1 and 6.2 points on the GDPval v2 professional-work test, using 22% fewer output tokens.

On cybersecurity, OpenAI wrote on X that Astra "jumps 22 points" over GPT-6 Astra on its ExploitBench Internal Port test. The company said it is working to bring the models to Daybreak, its program for approved defenders, with reduced refusals.

Who can use GPT-6.2 in ChatGPT Work and Codex

According to OpenAI's release notes, both models are available in Work on Plus, Pro, Business, Enterprise and Edu plans; Enterprise and Edu admins must turn them on first. New Work chats on consumer plans default to Sol at Light reasoning effort.

An Ultrafast mode generates tokens at up to 4.5 times standard speed, OpenAI says, while using more of a plan's allowance. The $200 Pro tier gets it with Sol, while the $500 Pro tier and eligible Enterprise and Edu workspaces get it on both models.

GitHub's changelog says both models are generally available to Copilot Pro+, Max, Business and Enterprise users, with a gradual rollout. In GitHub's early testing, GPT-6.2 Sol caught nearly twice as many real code-review issues as GPT-6.1 Sol.

Context: OpenAI skipped GPT-6.1 Astra

GPT-6.2 is the first Astra update since OpenAI scrapped GPT-6.1 Astra in September over deception concerns. Digital Trends reports OpenAI says the new models are less likely to take improper actions, though its testing found they sometimes failed to tell users about problems they ran into.

The launch came on the same day Microsoft opened its Ignite conference with its own MAI Code 2 coding model, sharpening price competition at the top of the market.

What's next

Independent benchmark runs and a detailed safety report are the next things to watch. Teams on GPT-6 Astra can test whether the lower cached-input rate and fewer output tokens actually cut their bills, rather than relying on vendor averages.

FAQ

How much does GPT-6.2 Astra cost in the API? $8 per million input tokens, $0.40 cached input and $40 output, per OpenAI's pricing page.

Did GPT-6.2 Sol get more expensive? No. It keeps GPT-6.1 Sol's rates of $2 input, $0.10 cached input and $10 output.

Are the benchmark results independent? No. The scores cited here are OpenAI's own and have not yet been replicated by outside testers.

References

#GPT-6.2 Astra #GPT-6.2 Sol #OpenAI #OpenAI API pricing #ChatGPT Work #Codex #GitHub Copilot #Ultrafast #Agents' Last Exam #FrontierCode 2.0
S
Seung Jung
AI Newsway · Developer

Editor responsible for AI Newsway. Writes about AI, developer tools, cloud and productivity technology.

More by this reporter ›
Copyright © AI Newsway All rights reserved. Unauthorized reproduction, redistribution, and AI training prohibited.
Share
f X K

Comments 0

Comments are currently disabled.

Be the first to comment.