█

AI/TLDR

Buy me a coffee
Buy me a coffee

GPT-6.2 Sol

OpenAI's balanced GPT-6.2 model for everyday coding and writing, close to Astra on code review standards at about a fifth of the cost.

IN SHORT

What is GPT-6.2 Sol?

GPT-6.2 Sol is OpenAI's efficient, general model for everyday work, released on November 17, 2026 alongside GPT-6.2 Astra. It succeeds GPT-6.1 Sol and keeps its token prices, while OpenAI says it delivers better results with fewer output tokens. It has a 1.05M-token context window and costs $2.00 per million input tokens and $10.00 per million output tokens. You can use it through OpenAI API.

Older version
GPT-6.1 Terra · 26 October 2026
Where to use it
OpenAI API
Updated Checked against 3 sourcesMarkdown version ↗
Cite this pageAI/TLDR, "GPT-6.2 Sol: specs, pricing and benchmarks", https://ai-tldr.dev/models/gpt-6-2-sol/ (updated 17 November 2026)
// FACTSas of 17 November 2026
Context window1.05Mmax output 128K[1]
Input price$2.00 / 1M tokenscached $0.10[2]
Output price$10.00 / 1M tokens[2]
LicenceProprietary · API only[1]
Released17 November 2026[3]
Knowledge cutoff1 June 2026[1]
  1. [1] deploymentsafety.openai.com GPT-6.2 Astra and Sol system card
  2. [2] developers.openai.com GPT-6.2 Sol model documentation
  3. [3] openai.com Introducing GPT-6.2 Astra and Sol (OpenAI, November 17, 2026)
GPT (Flagship / Thinking)API onlyAvailable
Released
17 Nov 2026
Context
1.05M
Input
$2.00 / 1M tokens
License
Proprietary

Overview

GPT-6.2 Sol is OpenAI's efficient, general model for everyday work, released on November 17, 2026 alongside GPT-6.2 Astra. It succeeds GPT-6.1 Sol and keeps its token prices, while OpenAI says it delivers better results with fewer output tokens. OpenAI positions it as the more affordable choice for daily coding and writing, with GPT-6.2 Astra reserved for the hardest projects.

In Cognition's FrontierCode 2.0, which checks whether code changes meet maintainers' standards for merging, GPT-6.2 Sol comes within 0.5 points of GPT-6.2 Astra at about one-fifth of the cost, with both at max reasoning effort. On OpenAI's LifeVal benchmark of everyday tasks, it outscores GPT-6 Astra at about one-seventh of the cost, and on OpenAI's internal GDPval v2 it improves 5.3 points over GPT-6.1 Sol. On Artificial Analysis's AA-3D human-preference test it scores 1,259 Elo.

In the API, gpt-6.2-sol accepts text and image input and returns text, with a 1,050,000-token context window (922,000 input tokens and up to 128,000 output tokens) and a knowledge cutoff of June 1, 2026. It runs on the Chat Completions, Responses and Batch endpoints. Pro 200 subscribers get Ultrafast mode for the first time with GPT-6.2 Sol; OpenAI rates the model Critical for cybersecurity under its Preparedness Framework, like GPT-6.2 Astra.

Released2026-11-17
LicenseProprietary
WeightsAPI only
Context1.05M
Max output128K
Knowledge cutoff2026-06-01
ModalitiesText, Vision
StatusAvailable

Pricing

Input$2.00 / 1M tokens
Cached input$0.10 / 1M tokens
Output$10.00 / 1M tokens

Same rates as GPT-6.1 Sol. GPT-6.2 Astra is $8.00 input and $40.00 output per 1M tokens.

Pricing source ↗

Strengths

  • Within 0.5 points of GPT-6.2 Astra on FrontierCode 2.0 at about a fifth of the cost
  • Same token prices as GPT-6.1 Sol: $2.00 input, $0.10 cached input, $10.00 output per 1M tokens
  • Up 5.3 points over GPT-6.1 Sol on OpenAI's GDPval v2
  • 1.05M-token context window with up to 128K output tokens
  • Ultrafast mode for Pro 200 subscribers and in the API

Best for

  • Reach for it for everyday coding in Codex where Astra-level review quality matters but cost adds up.
  • Reach for it for writing and document work that runs all day inside ChatGPT Work.
  • Reach for it for high-volume API agents that need long context at the Sol price.
  • Reach for it for everyday planning tasks such as comparing homes, schools or meal plans.

How to access

ProviderModel ID
OpenAI API ↗gpt-6.2-sol

GPT (Flagship / Thinking) — every version

The full lineage of the GPT (Flagship / Thinking) line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.

VersionReleasedContextLicense
GPT-6.2 Astracurrent2026-11-171.05MProprietary
GPT-6.2 Sol2026-11-171.05MProprietary
GPT-6.1 Terra2026-10-261.05MProprietary
GPT-6.1 Sol2026-09-291.05MProprietary
GPT-6 Sol2026-09-221.05MProprietary
GPT-6 Luna2026-09-221.05MProprietary
GPT-6 Astra2026-09-03—Proprietary
GPT-5.6 Sol2026-07-091.05MProprietary
GPT-5.6 Terra2026-07-091.05MProprietary
GPT-5.6 Luna2026-07-091.05MProprietary
GPT-5.52026-04-231.05MProprietary
GPT-5.42026-03-05—Proprietary
GPT-5.22025-12-11—Proprietary
GPT-5.12025-11-12—Proprietary
GPT-52025-08-07—Proprietary
GPT-4o2024-05-13—Proprietary

FAQ

How much does GPT-6.2 Sol cost?

OpenAI lists GPT-6.2 Sol at $2.00 per 1M input tokens, $0.10 per 1M cached input tokens and $10.00 per 1M output tokens, the same rates as GPT-6.1 Sol. Cache writes cost $2.50 per 1M tokens, and GPT-6.2 Astra costs four times as much per token.

Should I use GPT-6.2 Sol or GPT-6.2 Astra?

OpenAI recommends GPT-6.2 Sol for everyday coding and writing and GPT-6.2 Astra for the hardest projects. On FrontierCode 2.0, GPT-6.2 Sol comes within 0.5 points of GPT-6.2 Astra at about a fifth of the cost, so Sol covers most coding work while Astra leads on long professional and scientific tasks.

Who can use GPT-6.2 Sol?

GPT-6.2 Sol launched on November 17, 2026 for Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and in the API as gpt-6.2-sol. Pro 200 subscribers get Ultrafast mode with GPT-6.2 Sol for the first time, and Ultrafast is available for it in the API at launch.