GPT-6.1 Sol

// Overview

GPT-6.1 Sol is OpenAI's upgrade to GPT-6 Sol, announced at OpenAI DevDay on September 29, 2026, one week after GPT-6 Sol itself shipped. OpenAI's API documentation describes it as offering "near-Astra performance at a lower cost" for complex coding, computer use and professional work. It was released on its own: the planned GPT-6.1 Astra was cancelled the day before over safety test results, so GPT-6 Astra remains the top tier.

The model 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 April 30, 2026. It is served on the Chat Completions, Responses and Batch endpoints, and supports streaming, structured outputs, function calling, file search, web search, prompt caching and hosted tools such as code interpreter, computer use and hosted shell. Reasoning effort can be set to low, medium (the default), high, xhigh or max.

The list price matches GPT-6 Sol at $2.00 per 1M input tokens and $10.00 per 1M output tokens, a fifth of GPT-6 Astra's $10.00 and $50.00. Cached input drops to $0.10 per 1M tokens, half of GPT-6 Sol's $0.20. The Next Web reports that GPT-6.1 Sol beats GPT-6 Sol by 6.4 points on DeepSWE v1.1, comes within 2.1 points of Astra on OSWorld 2.0, and cuts factual errors at low reasoning effort from 11.4% to 7.7%. At launch it was available in the API as gpt-6.1-sol and to Plus, Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex, but not yet in the main ChatGPT chat.

// Pricing

Same input and output rate as GPT-6 Sol, with cached input halved from $0.20. GPT-6 Astra is $10.00 input and $50.00 output per 1M tokens.

Pricing source ↗

// Strengths

• Near-Astra agentic coding and computer use at a fifth of GPT-6 Astra's token price
• Cached input at $0.10 per 1M tokens, half the GPT-6 Sol rate
• 1.05M-token context window with up to 128K output tokens
• Hosted tools including computer use, code interpreter and hosted shell on the Responses API
• Reasoning effort configurable from low through max

// Best for

• Reach for it for agentic coding in Codex when GPT-6 Astra's price is hard to justify.
• Reach for it for computer-use agents that operate browsers and desktop apps through the Responses API.
• Reach for it for long-context document work that needs Astra-like reliability at a lower cost.
• Reach for it for agent loops that re-send the same long prompt, where the cheaper cached input adds up.

// How to access

// 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.