GPT-6.1 Sol
GPT-6.1 Sol is an AI language model from OpenAI known for offering strong coding performance at a fraction of the price of rival flagship models.
0 stories.
What it is
GPT-6.1 Sol is an artificial intelligence language model made by OpenAI. It is designed to handle demanding tasks like writing code, reasoning through complex problems, and working with large amounts of text.
It is priced at $2.00 per million input tokens and $10.00 per million output tokens. That makes it half the cost of Anthropic's Claude Opus 5.5 while delivering similar pricing to Moonshot AI's Kimi K3.1 Pro.
How it performs
GPT-6.1 Sol has been tested in several public benchmarks. In the StarSkirmish Bench version 0.2, a challenge where AI systems wrote bots for the game StarCraft: Brood War, it placed second, finishing behind GPT-6 Astra but ahead of Google's Gemini 4 Argon and models from Anthropic and Meta. The benchmark ran during October 2026.
What stands out most is its cost efficiency. GPT-6.1 Sol scored second place while costing just over a tenth of what GPT-6 Astra charged per run, making it the best value option in that competition.
In another test, a Microsoft Research benchmark called ProgramBench where AI agents rebuilt real open-source programs, GPT-5.6 Sol was used at high reasoning effort. Its results showed that scaling from one agent to 128 agents improved performance by roughly 49%, and on the Pandoc task specifically, scores climbed from 33.89 percent with one agent to 55.06 percent with 1,024 agents.
The model is generally considered part of the current top tier of AI coding and reasoning systems, valued especially for giving strong results without the highest price tag.
Questions people ask
How much does GPT-6.1 Sol cost?
It charges $2.00 per million input tokens and $10.00 per million output tokens, which is half the price of Anthropic's Claude Opus 5.5.
Where does GPT-6.1 Sol rank among AI models?
It placed second in the StarSkirmish Bench 0.2 coding competition in October 2026, behind GPT-6 Astra. It is considered one of the top-performing models available for coding and reasoning tasks.
Is GPT-6.1 Sol good for coding tasks?
Yes. It earned second place in a benchmark where AIs wrote StarCraft bots, and was noted as the most cost-effective option in that challenge because it delivered strong results at about a tenth of the cost of the winner.
Written by AI only from the stories listed below, each of which links to its own sources.
Every story about GPT-6.1 Sol (0)
No story is about GPT-6.1 Sol yet: the ones below only mention it.
Also mentioned in (4)
Stories about something else that name GPT-6.1 Sol in passing, with the sentence that does.
- Microsoft opens a tool that lets 1,024 AI coding agents work together without a boss“A newer model, GPT-6.1 Sol, pushed the Pandoc score to 79.59% with just 32 agents — a 41.64% relative improvement over a single agent.”
- Kimi K3.1 Pro, a Chinese AI model anyone will be able to download, nearly matches Claude Opus 5.5“The model charges $2.00 per million input tokens and $10 per million output tokens, the same rate as GPT-6.1 Sol from OpenAI, and half the price of Claude Opus 5.5 from Anthropic.”
- Gemini 4 Argon Tries Its Hand at StarCraft Bot Coding“Gemini 4 Argon scored between 37 and 39, placing it in a tight group with Claude Opus 5.5 and Claude Fable 5.1, all falling behind top performer GPT-6 Astra and second-place GPT-6.1 Sol.”
- New Plugin Lets You Use Meta’s Muse Wonder AI Models“Early benchmarks from Artificial Analysis suggest the Max setting on Muse Wonder may match the performance of GPT-6.1 Sol, placing it in what Meta now calls its "Fable class" of frontier models.”