GPT-6.2 Astra And Sol: What Their Arrival Could Change At Work
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: GPT-6.2 Astra And Sol: What Their Arrival Could Change At Work on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get smart everyday buys delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

OpenAI released GPT-6.2 Astra and Sol on November 17, 2026, alongside Astra for Law 6.2. Early Artificial Analysis results place Astra near the top of its intelligence index and report lower costs on one benchmark task than several rivals, but coding-agent results did not improve; OpenAI rates both new models Critical for cybersecurity.

OpenAI released GPT-6.2 Astra and GPT-6.2 Sol on November 17, alongside Astra for Law 6.2, positioning Astra for demanding projects and Sol for lower-cost everyday work and agents. Early tests from Artificial Analysis put Astra among the strongest model families, while the release also comes with a significant caveat: OpenAI’s safety report rates both models Critical for cybersecurity.

Artificial Analysis’s Intelligence Index, accessed on launch day, gives Astra Max a score of 62 and Sol Max a score of 57. The index’s leading model family, Anthropic’s Claude Fable 5.2, scores 65 in the summary comparison. Astra’s unrounded score is 62.14, just below Fable 5.2 at 62.22 in the cited configuration; Astra’s model page lists it third among 91 configurations. The results indicate a close race near the top, not a clear lead for OpenAI.

Costs vary by model and configuration. Artificial Analysis estimates one index task at $4.00 for Astra Max and $1.01 for Sol Max, compared with $26.54 for Claude Fable 5.2 Max and $5.48 for Gemini 4 Carbon High. These are costs for that benchmark task, not universal prices for workplace use. OpenAI’s listed API rates for standard short-context requests are $8 per million input tokens and $40 per million output tokens for Astra, and $2 and $10 for Sol. Astra’s input and output rates are 20% below those of GPT-6 Astra, the previous model, while Sol retains the same listed rates.

Both models accept text and images, have a 1,050,000-token context window, and can return up to 128,000 output tokens. OpenAI lists Astra for Law 6.2 as a legal model configuration at the same standard rates as Astra; the documentation reviewed does not explain its differences from the general model. Artificial Analysis has not assigned it an Intelligence Index score, though it has tested the legal configuration on one legal benchmark.

At a glance
reportWhen: Released November 17, 2026; early evalu…
The developmentOpenAI released GPT-6.2 Astra and Sol, plus a legal configuration of Astra, with independent benchmark results and the company’s safety report published the same day.

GPT-6.2 Astra, Sol and Astra for Law

Three models, three jobs

What launched, what it costs and how it scores in independent tests. Astra and Sol at their Max setting; the better value in each row is outlined.

AstraSolFor Law Price in/out$8/$40$2/$10$8/$40 Index score6257– Cost per task$4.00$1.01– Minutes per task10.47.9– Legal test15.0%11.1%20.8% Both rated Critical for cyber risk
  1. Three releases, one dayOn November 17, OpenAI released GPT-6.2 Astra for the hardest projects and GPT-6.2 Sol for everyday work. Astra for Law 6.2 is a legal configuration at Astra’s price.
  2. The independent indexAstra scores 62, second among model families behind Claude Fable 5.2 at 65. Sol scores 57. One index task costs $4.00 and $1.01.
  3. Speed and a legal testAstra needs 10.4 minutes per task, Sol 7.9. On the Harvey LAB-AA legal test, Astra for Law passes 20.8% of tasks.
  4. OpenAI’s own safety ratingBoth models meet OpenAI’s Critical threshold for cybersecurity. Higher-risk cyber help now needs extra vetting.

Source: Artificial Analysis, model pages for GPT-6.2 Astra, GPT-6.2 Sol and Astra for Law 6.2, accessed Nov 17, 2026 (Intelligence Index, cost and time per task, Harvey LAB-AA v1.1); OpenAI, API pricing page and GPT-6.2 System Card, Nov 17, 2026. Prices per million input/output tokens. Redrawn by Thorsten Meyer AI.

Graphic: Powered by Thorsten Meyer AI

Potential gains for workplace AI

The release gives organizations a choice between a higher-capability model and a less expensive option, which could affect the cost of deploying AI for coding, document work and automated agents. Artificial Analysis’s estimate that Sol completes its index task for $1.01, compared with $4 for Astra, offers a concrete example of the trade-off, though results on one test cannot establish costs or quality across a company’s workload.

For buyers, the headline performance numbers are only part of the decision. The Critical cybersecurity ratings, reported hallucination rates of 59% for Astra and 55% for Sol at maximum effort, and the absence of improvement in the Coding Agent Index all bear on whether either model is suitable for a particular workflow. Organizations will need to weigh capability, price, oversight and security controls rather than treating the launch as a general productivity guarantee.

Pricing also depends on how the API is used. Requests exceeding 272,000 input tokens incur higher rates across the request; Batch costs half the standard rate, while Fast costs twice as much. Ultrafast is listed at three times standard through February 28, 2027, then six times standard from March 1. The actual bill will depend on model, prompt length, processing tier and usage.

De'Longhi Stilosa Compact Manual Espresso Machine with Milk Frother

De'Longhi Stilosa Compact Manual Espresso Machine with Milk Frother

As an affiliate, we earn on qualifying purchases.

How OpenAI positioned the models

OpenAI’s developer guide describes Astra as the option for the most demanding tasks and Sol as a lower-cost choice for complex work. The launch therefore presents a two-model lineup rather than a single replacement: a premium model and a smaller-cost alternative intended for everyday work and agents at scale. OpenAI’s launch post was not accessible to the source author, so the models’ inclusion in specific ChatGPT plans could not be confirmed.

The release also changes some API behavior. OpenAI says that, starting with GPT-6.2, editing conversation history mid-conversation removes the model’s earlier reasoning. The company warns that answer quality can decline and that the model may spend more tokens repeating earlier work. API requests that set reasoning effort to “none” are rejected.

OpenAI’s safety report was published on the same day as the models. It identifies improvements as well as regressions, according to the source report. One stated capability threshold is also notable: Astra is the first model OpenAI has released externally that meets the company’s High threshold for AI self-improvement. That designation is OpenAI’s own assessment, not an independent finding presented in the available material.

Questions on safety and deployment

The available results do not establish how Astra and Sol will perform across specific workplace tasks or production systems. Artificial Analysis’s task-cost estimates reflect its own benchmark and configurations; they should not be read as a direct forecast of an organization’s API bill. The source material also does not provide enough detail to compare every test result or explain the differences between Astra for Law and standard Astra.

OpenAI rates both models at its Critical cybersecurity threshold, the higher of the two risk thresholds it uses, and says more permissive cyber access requires additional vetting. The materials provided do not specify the full access rules or how those controls will be applied across products. OpenAI’s report also lists regressions alongside improvements, but the available account does not detail every regression.

It is also unclear which ChatGPT subscriptions include the models, since OpenAI’s launch post could not be reviewed by the source author. Rate limits depend on account usage tier, and the reviewed pricing information does not resolve how costs will compare for individual customers’ workloads. These points may become clearer as product documentation and independent evaluations expand.

Documentation and testing ahead

Users and developers can compare the models through the API pricing and model documentation, while organizations with approved accounts can check availability, usage limits and data-handling options. OpenAI lists Zero Data Retention for approved organizations and US and EU data residency with Standard, Fast and Batch processing; Ultrafast is listed with US data residency only. The company says promotional pricing for Astra for Law is available at least through December 31, 2026.

Further independent testing is likely to clarify how the models compare across workplace tasks, including coding agents and legal work. Artificial Analysis has not yet scored Astra for Law on its Intelligence Index. For now, the launch establishes the models’ stated roles, listed API prices and initial evaluation results, but not their real-world value for any particular organization.

Key Questions

What did OpenAI release on November 17, 2026?

OpenAI released GPT-6.2 Astra and GPT-6.2 Sol, as well as Astra for Law 6.2, a legal model configuration.

How do Astra and Sol differ?

OpenAI positions Astra for the most demanding tasks and Sol for complex work at lower cost. Standard short-context API rates are $8 per million input tokens and $40 per million output tokens for Astra, versus $2 and $10 for Sol.

Do early tests show GPT-6.2 leading all competitors?

No. Artificial Analysis places Astra near the top of its Intelligence Index, but its unrounded score of 62.14 is slightly below Claude Fable 5.2’s 62.22 in the cited configurations. The evaluator also reported no improvement in its Coding Agent Index.

What does OpenAI’s cybersecurity rating mean?

OpenAI’s safety report rates both Astra and Sol at its Critical threshold for cybersecurity, the higher of the two risk thresholds it uses. The source says more permissive cyber access requires extra vetting, but does not provide the full access rules.

Are the models included in ChatGPT plans?

The available source material does not confirm which ChatGPT plans include Astra or Sol. OpenAI’s launch post was not accessible to the source author, so product availability by subscription remains unclear.

Source: ThorstenMeyerAI.com

This content is for general information only and is not financial, tax or legal advice. Consult a qualified professional for decisions about your money.
You May Also Like

Moving Abroad Instead of Equity Release: The Guide Most Homeowners Needed Sooner

Will choosing to move abroad instead of releasing equity preserve your assets and lifestyle—discover the key factors before making your decision.

Six AI Milestones Expected To Happen In 2026

Predictions for six major AI advancements set to occur in 2026, including breakthroughs in natural language processing, robotics, and ethical frameworks.

Qwen4 Max Cuts The Cost To A Third, Yet The Leaders Keep Their Lead

Artificial Analysis scores Alibaba’s Qwen4 Max at 43 and $3.27 per index task, below leading models but cheaper than its predecessor.