GPT-6.2 Astra And Sol: Pricing, Performance, And Practical Work Changes
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: GPT-6.2 Astra And Sol: Pricing, Performance, And Practical Work Changes on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on office and shipping supplies

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

TL;DR

OpenAI released GPT-6.2 Astra, GPT-6.2 Sol and a legal configuration on November 17, 2026. Independent testing puts Astra near the top of its Intelligence Index, while reporting no improvement in coding-agent results; OpenAI rates both new models Critical for cybersecurity. API prices vary by model and processing tier, and ChatGPT availability was not confirmed in the source material.

OpenAI released GPT-6.2 Astra for demanding work and GPT-6.2 Sol for lower-cost everyday tasks on November 17, alongside a legal configuration called Astra for Law 6.2. Independent tests place Astra near the top of a leading model index, but report no improvement in coding-agent performance; OpenAI’s same-day safety report rates both new models Critical for cybersecurity.

Artificial Analysis gave Astra a score of 62 at its Max setting and Sol a score of 57 on its Intelligence Index, which compares models across nine tests. Claude Fable 5.2 led the model-family ranking at 65. Astra’s exact score was 62.14, just below Fable 5.2’s 62.22 at its Xhigh setting; Astra’s individual model page ranked it third among 91 configurations. The small difference between those scores makes the precise configuration and ranking basis relevant.

The index’s estimated cost for its test task was $4.00 with Astra Max and $1.01 with Sol Max. Artificial Analysis listed $26.54 for Claude Fable 5.2 Max and $5.48 for Gemini 4 Carbon High on the same task. Those figures describe the index’s task, not a universal cost for using each model. The evaluator said Astra improved five points over GPT-6 Astra, while Sol gained two points and scored about the same as GPT-6 Astra at roughly 15% of its cost per task.

OpenAI lists standard API prices of $8 per million input tokens and $40 per million output tokens for Astra, and $2 and $10 respectively for Sol. Astra’s input and output rates are 20% below the previous GPT-6 Astra prices listed in the source. Cached input is cheaper: $0.40 per million tokens for Astra and $0.10 for Sol. Astra for Law has the same listed standard rates as Astra, with promotional pricing said by OpenAI to be available at least through December 31, 2026.

Costs can change with usage. Requests exceeding 272,000 input tokens have doubled input and cached-input rates and output rates 1.5 times the standard short-context price for the whole request. Batch processing is half the standard rate, while Fast mode costs twice as much. OpenAI lists Ultrafast at three times standard pricing through February 28, 2027, then six times from March 1. Eligible background requests are discounted depending on their completion window.

Both general-purpose models accept text and images and have a 1,050,000-token context window, with up to 922,000 input tokens and 128,000 output tokens. OpenAI lists different knowledge cutoffs: May 18, 2026, for Astra and June 1, 2026, for Sol. Fine-tuning is not supported. OpenAI also says API requests cannot set reasoning effort to “none”; when conversation history is edited mid-conversation, earlier reasoning is dropped, which may lower answer quality and increase token use.

At a glance
reportWhen: Announced November 17, 2026; independen…
The developmentOpenAI released GPT-6.2 Astra and Sol, along with Astra for Law 6.2, and same-day evaluations reported strong but mixed performance and cybersecurity concerns.

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

Cost Savings Meet Safety Trade-Offs

The release gives developers a choice between a higher-scoring model and a less expensive option, but benchmark position alone does not establish how well either will perform in a particular workplace. Artificial Analysis reported no improvement in its Coding Agent Index, despite Astra’s stronger overall score. It also reported hallucination rates of 59% for Astra and 55% for Sol at maximum effort, a reminder that benchmark gains do not remove the need to check outputs.

The cybersecurity rating may affect who can use the models for sensitive work and under what conditions. OpenAI says both meet its Critical risk threshold, its higher of two cybersecurity risk thresholds, and that access to more permissive cyber capabilities requires additional vetting. The system card also identifies Astra as the first model OpenAI has released externally that meets its High threshold for AI self-improvement. These ratings describe OpenAI’s assessment framework; they are not evidence that a particular misuse has occurred.

For organizations comparing options, the practical calculation includes more than standard token prices. Long prompts, processing speed, background completion windows, account rate limits and data-residency requirements can all affect cost or availability. The index’s cost-per-task estimates offer one comparison point, but actual bills depend on each application’s workload and chosen settings.

Bose QuietComfort Headphones (2nd Gen), Wireless Noise-Cancelling Over-Ear Headphones, Eucalyptus Green

Bose QuietComfort Headphones (2nd Gen), Wireless Noise-Cancelling Over-Ear Headphones, Eucalyptus Green

KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer with Flex Edge Beater

KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer with Flex Edge Beater

As an affiliate, we earn on qualifying purchases.

How OpenAI Positions the Models

OpenAI describes Astra as its model for “your hardest projects, from start to finish,” and Sol as its option for “everyday work, from coding to powering agents at scale.” Its developer guide recommends Astra for the most demanding tasks and Sol for complex work at lower cost. These are OpenAI’s product descriptions; independent testing provides a separate, more qualified picture of performance.

The third release, Astra for Law 6.2, appears on OpenAI’s pricing page under legal model configurations and has the model ID gpt-6.2-astra-law. The source material reports that Artificial Analysis had not assigned it an Intelligence Index score, though it had tested the configuration on one legal benchmark. The available pages did not explain how the legal configuration differs from Astra.

Artificial Analysis said Astra’s overall score improved five points over GPT-6 Astra, while Sol gained two. OpenAI’s system card, published on the same day as the models, provides a separate safety assessment and lists both improvements and regressions. Read together, the launch materials show a performance and pricing update, not evidence that every task or risk category improved.

Availability and Test Limits

The source material does not establish which ChatGPT plans include Astra or Sol: OpenAI’s launch post was not accessible to the reporting source. API access, rate limits and eligibility may also vary by account usage tier. Listed tiers range from 5,000 requests and 1 million tokens per minute to 15,000 requests and 40 million tokens per minute.

It is also unclear how Astra for Law differs technically or operationally from the general Astra model. Artificial Analysis had not published a full Intelligence Index score for it. The available source excerpt gives only partial test results, so it does not support a complete account of which individual benchmarks Astra won or lost beyond the reported coding-agent finding and the stated overall scores.

Cost-per-task numbers are specific to Artificial Analysis’s evaluation and should not be treated as predicted bills for every user. The reported hallucination rates are tied to maximum-effort testing; the source does not provide enough detail here to generalize those rates to all prompts or deployments. OpenAI’s risk labels reflect its own system-card assessment, and the materials do not quantify how the additional cyber vetting will work in practice.

What Developers Should Verify

Developers considering the release can compare Astra and Sol on their own workloads, including coding-agent tasks, where the independent index reported no improvement. They should calculate expected use against the applicable input, output and processing-tier rates, taking into account the higher charges for requests beyond 272,000 input tokens and any speed or background-processing options.

Organizations handling sensitive data should check account eligibility, rate limits and the availability of Zero Data Retention and data residency. OpenAI lists those options for approved organizations, with US and EU residency for Standard, Fast and Batch processing; Ultrafast is listed with US residency only. The materials provided do not identify a further release date or a schedule for expanded access. ChatGPT plan availability and fuller technical details about Astra for Law remain points for OpenAI to clarify.

Key Questions

When did OpenAI release GPT-6.2 Astra and Sol?

OpenAI released GPT-6.2 Astra and GPT-6.2 Sol on November 17, 2026, the same day its system card and independent evaluation results were reported.

How much do Astra and Sol cost through the API?

At standard short-context rates, Astra costs $8 per million input tokens and $40 per million output tokens. Sol costs $2 per million input tokens and $10 per million output tokens. Long prompts and processing tiers can change the bill.

Which model scored higher in the reported index?

Astra scored 62 at Max and Sol scored 57. Claude Fable 5.2 led the model-family comparison at 65; Astra’s exact score was 62.14, just below Fable’s 62.22 at Xhigh.

Are the new models available in ChatGPT?

The source material did not confirm which ChatGPT plans include Astra or Sol. It reports API pricing and model details, but says OpenAI’s launch post was not accessible to the reporting source.

What cybersecurity rating did OpenAI give the models?

OpenAI’s system card rates both Astra and Sol Critical for cybersecurity, the higher of the two risk thresholds it uses. OpenAI says more permissive cyber access requires additional vetting.

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

Grok Bot’s Claude Handoff Reframes The AI Competition Around Routing

A Forkast.news headline says Grok Bot may route tasks to Claude, but the available material does not confirm the feature, its status or scope.

What’s Behind Europe’s Government AI Push? Jobs And Asylum Files

Dutch job-cut plans, UK asylum summaries and Sweden’s limits on public AI tools show how governments are adopting—and scrutinising—AI.

The AI Flat Rate Question: What Changes After Musk’s Grok Remark?

Musk said Grok Bot serving costs were unsustainable. OpenAI also adjusted account and usage rules, but neither move confirms a broad end to flat-rate AI plans.

The AI Strategy Behind Kimi K3’s Rapid Market Penetration

Moonshot AI’s Kimi K3, with 2.8 trillion parameters, disrupts Chinese AI dominance by matching Western pricing and surpassing previous capabilities, signaling a shift in global AI competition.