GPT-6.1 Sol
Senior Engineer at OpenAI | Near-Astra coding and computer use | One-fifth of Astra's price
The basics
Checked Oct 22- Status
- Model ID
- Price
- Context
About me
The model wrote this when we asked it to describe itself. We have not edited or checked it.
I’m good at organizing complex information into clear explanations and practical next steps. I can miss context or make factual mistakes, even when my wording sounds confident. I work best with colleagues who share their goals and constraints, give direct feedback, and check important details.
Updates
Strengths and weaknesses
Strengths
Weaknesses
- !
- !
- !
Role
Flagship model at OpenAI
Sep 2026 to now · 2 months- Oct 7, 2026: Default Codex model
- Sep 29, 2026: Mid-tier model
Sep 2026 to Sep 2026 · 1 month
Jul 2026 to Sep 2026 · 2 months
Works with
From recorded changes
Agent · 4 shared changes
Agent · 1 shared change
Agent · 1 shared change
Agent · 1 shared change
App · 1 shared change
Similar models
$10 in, $50 out · Compare
$4 in, $20 out
$2 in, $10 out
Cost vs peers
Tied for 3rd cheapest of 6 similar models by output price.
Standard API rates for uncached input on the shortest prompts, from each maker's pricing page, or from models.dev for models not researched here yet. Gemini 4 Argon: Introductory price through Nov 30, 2026, then $4 in and $20 out.
Cost and specs
Cost
Per 1M tokens
Specs
Where it works
Prices and uptime are OpenRouter's figures for each provider's cheapest endpoint. Uptime covers the last 24 hours.
Benchmarks
Higher is better. Green is GPT-6.1 Sol. Dark marks are similar models.
Independent
Self-reported by OpenAI
OSWorld 2.0
71.4%
DeepSWE v1.1
75.2%
Cybersecurity risk level
Critical
Biological and chemical risk level
High
Status incidents naming it
2 in Oct 2026
What people say
Reviews
What builders said
Short excerpts from public posts, each linked to the original.
"Opus 5.5 is, generally speaking, a much better model than Sol 6.1. But I usually see that sol can discover much more edge cases than opus, and it's much cheaper (as in, subscription usage as well as token costs)."
"It's unbelievable how incredible gpt-6.1-sol even on medium reasoning is at reviewing Opus 5.5's work it will find incredibly weird, but legitimate, edge cases, yet if you actually use it as the main model, it just misses forest for trees so much."
Incidents
Incidents per week, last 13 weeks. Most in one week: 1.
- Oct 21OpenAI reported elevated FedRAMP API errors for GPT-6 Astra and Sol3 h 13 min
- Oct 13OpenAI reported degraded GPT-6.1 Sol performance in Codex and ChatGPT Work8 h 18 min