OpenAI · llm model

GPT-6.1 Sol

Release Ruler for GPT-6.1 Sol: released Sep 29, 2026, $2 per 1M input tokens, proprietary. Each thin stem on the ruler is another model we track.

GPT-6.1 Sol is an OpenAI language model released Sep 29, 2026; OpenAI's newest language model in our dataset is GPT-6.1 Terra. Proprietary, API access only. OpenAI's near-flagship GPT-6.1 model: OpenAI says it performs close to GPT-6 Astra at a fifth of Astra's price. Verified against vendor sources on Oct 26, 2026.

Reported Last verified · All models · JSON · Markdown

  • ReleasedSep 29, 2026
  • Context window1.05M1,050,000 tokens
  • Max output128K128,000 tokens
  • LicenceProprietaryProprietary (API)
InTextImageOutText

Key facts

GPT-6.1 Sol: key facts with sources
FactValueSource
VendorOpenAIn/a
ReleasedSep 29, 2026source
Statusgenerally availablen/a
API idgpt-6.1-solsource
Context window1,050,000 tokens (1.05M)source
Max output128,000 tokenssource
Inputtext, imagesource
Outputtextsource
LicenceProprietary (API)source
Pricing$2 in / $10 out per 1M tokenssource

Where GPT-6.1 Sol sits in our data

How GPT-6.1 Sol differs from GPT-6 Sol

Read from the recorded facts of the two models, not from a head-to-head test. We have not run either.

  • Listed price is the same as GPT-6 Sol: $2 in and $10 out (per 1M tokens).
  • GPT-6.1 Sol lists a 1.05M-token context window; we have none recorded for GPT-6 Sol.
  • Both accept text, image as input.
  • GPT-6.1 Sol came out 7 days after GPT-6 Sol.
  • Facts for the other model: GPT-6 Sol.

What the sources say

These are claims from the linked pages, not our measurements.

  • Vendor claim (OpenAI): near-Astra performance at one fifth of Astra's price. We have not independently measured this.

Notes

  • At launch it was available in ChatGPT Work and Codex, not yet in ChatGPT Chat. Knowledge cutoff listed: Apr 30, 2026.

What we tested

Nothing yet. We have not run GPT-6.1 Sol ourselves, so every fact on this page is Reported from the sources below. When we do test it, the scope will be stated here and in our test logs. See methodology.

Sources

More from OpenAI

All 88 models