MA
Google just went FRONTIER (Gemini 4 Argon)Matthew Berman14 Oct 2026
OpenAI model, described as great. It scores 71.4 on OS World versus Argon's 69. It was added to the creator's own benchmark chart.
GS
GPT-6.1 Sol
by OpenAI
OpenAI's GPT-6.1 Sol: near-GPT-6 Astra performance on coding and agent tasks at one-fifth of Astra's API price.
Updated EU status checked Last shown
GPT-6.1 Sol at a glance
- What is GPT-6.1 Sol?
- GPT-6.1 Sol is a model from OpenAI. OpenAI's GPT-6.1 Sol: near-GPT-6 Astra performance on coding and agent tasks at one-fifth of Astra's API price.
- Who makes GPT-6.1 Sol?
- GPT-6.1 Sol is made by OpenAI.
- Does GPT-6.1 Sol work in the EU?
- Available in the EU. The official page lists no regional restrictions and gives availability by plan (Plus, Pro, Business, Enterprise, Edu) and via the API. OpenAI's privacy pages mention GDPR support, a DPA, data residency for ChatGPT and API, and an EU DSA transparency report. Last verified 14 Nov 2026.
- How much does GPT-6.1 Sol cost?
- API pricing. API: $2/M input, $0.10/M cached input, $10/M output. Included in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu plans. Not yet in ChatGPT Chat. Licence: Proprietary.
- Where was GPT-6.1 Sol shown?
- In 7 creator videos by Matthew Berman, Matt Wolfe, AI Search. Most recently by Matthew Berman on 14 Oct 2026, at 9:56 in “Google just went FRONTIER (Gemini 4 Argon)”.
Sources: creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
About
GPT-6.1 Sol is an OpenAI model that upgrades GPT-6 Sol. OpenAI positions it between GPT-6 Astra (its most capable model) and GPT-6 Luna (its low-cost model). It is aimed at agentic coding, computer use and professional work. OpenAI reports it matches Astra on DeepSWE v1.1 and comes within about 2 points on OSWorld 2.0 (offline set, max reasoning effort), at a fraction of the cost per task.
OpenAI also reports gains over GPT-6 Sol on document question answering (GDP.pdf), business workflow automation (AutomationBench) and scientific workflows (Terminal-Bench Science 0.1). On a factuality test of prompts that had previously triggered errors, the factual-error rate at low reasoning effort fell from 11.4% to 7.7%. Astra still scores highest on Terminal-Bench Science (68.1%), and OpenAI recommends it for the hardest research tasks.
Access is through ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu users, and through the OpenAI API as gpt-6.1-sol. It is not yet available in ChatGPT Chat. API pricing is $2 per million input tokens, $0.10 per million cached input tokens and $10 per million output tokens. The name in this entry's alias list ("Ultra Fast" / "Ultrafast mode") refers, according to a YouTube creator, to GPT-6 Astra running on Cerebras hardware. That reportedly runs about eight times faster and costs about six times more. This is not confirmed in the provided official page text, so treat it as unverified.
Key features
- Near-Astra results on agentic coding, computer use and professional work at one-fifth of Astra's API price
- API pricing of $2 input, $0.10 cached input and $10 output per million tokens
- Available in ChatGPT Work and Codex, and via the API as gpt-6.1-sol
- Higher factual accuracy than GPT-6 Sol on difficult prompts
- Improved alignment evaluation results over GPT-6 Sol
9 moments
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
MA
Meta just dropped something insane... (Muse Wonder)Matthew Berman13 Oct 2026
The creator's favorite model, priced at $2/$10 per million tokens. It is cheapest per task at 72 cents and is the best value given quota and intelligence.
Everything has been solved
The creator tasked GPT-6.1 Soul with recreating Super Mario World from scratch for the Mod Retro handheld. It has been running in a loop for over 4 days and may finish in a few more. He says cheaper, more efficient models like this make recreating software easy and that he struggles to burn through his tokens.
MA
AI News: Haiku-5.5, Visual ChatGPT, Grok's New Trick, and more!Matt Wolfe9 Oct 2026
Ultra-fast mode for GPT-6.1 Soul in the API, Codex and ChatGPT Work. It is 8x faster at 6x the standard price ($2/$10 per million tokens standard) and requires the $500 Pro plan.
AI
Gemini 4, GPT 6.1, Plasma conjecture, Napoleon code solved, OpenAI Dots, Claude Sonnet 5.5: AI NEWSAI Search4 Oct 2026
An upgrade of GPT6 Sol, claimed to be near Astra intelligence for a fifth of the price. It is one point behind Astra on the Artificial Analysis index and is the most cost-efficient frontier model. It is available to paid plans in GPT work and Codex. Creator also notes the Pro plan usage cut and a new $500 plan.
MA
AI News: Dots, GPT-6.1 Sol, Sonnet 5.5, Gemini 4, and everything you need to knowMatt Wolfe2 Oct 2026
Coming-soon mode that reportedly runs on Cerebras chips, generating 300 tokens per second (about 8x faster). It costs 6x the price and requires the $500/month plan.
MA
AI News: Dots, GPT-6.1 Sol, Sonnet 5.5, Gemini 4, and everything you need to knowMatt Wolfe2 Oct 2026
A cheaper model almost as good as GPT-6 Astra, priced at $2/$10 per million tokens versus $10/$50. Scores 52 on Artificial Analysis at about 72 cents per task. Available on paid plans in work mode and Codex, not yet in chat.
OpenAI COOKED
GPT-6 Astra running on Cerebras chips, eight times faster than on Nvidia GPUs but about six times more expensive. The creator burned $1,000 in tokens in an hour and a half and showed a side-by-side demo where Ultrafast launched a rocket before standard finished thinking.
MA
OpenAI COOKEDMatthew Berman30 Sept 2026
A smaller, cheaper, faster model that is nearly as good as GPT-6 Astra, priced at $2/M input, $10/M output and $0.10/M cached input. Benchmarks shown include a developer-preference benchmark, GDPval-style PDF tasks and OSWorld, where it matches or is near Astra at a fraction of the price.
Keep exploring
Similar AI tools
Closest neighbours by meaning, not just by category.
GS
GPT-6 Sol
Model · OpenAI
OpenAI's mid-tier GPT-6 model, positioned below GPT-6 Astra and above Luna, priced at $2/$10 per million tokens.
Available in EULanguage Models
Release date n/aPaid
GS
GPT-5.6 Sol
Model · OpenAI
OpenAI's GPT-5.6 Sol frontier model, with an Ultrafast API tier on Cerebras reaching up to 750 output tokens per second.
Limited in EULanguage Models
Release date n/aAPI
GA
GPT-6 Astra
Model · OpenAI
OpenAI's GPT-6 Astra, a frontier language model used for reasoning, agentic work and weight interpretation research.
UnverifiedLanguage Models
Release date n/aAPI
GT
GPT-6.1 Terra
Model · OpenAI
OpenAI's mid-tier GPT-6.1 model for computer use, visual reasoning and coding, priced at $1/$5 per million tokens.
Available in EULanguage Models
Release date n/aPaid
GP
GPT-5.4
Model · OpenAI
OpenAI frontier language model used for coding, reasoning and agentic tasks, available via ChatGPT and the API.
Available in EULanguage Models
Release date n/aAPI
GP
GPT-6.1
Model · OpenAI
OpenAI language model in the GPT-6.1 family; the Soul Max variant is noted by reviewers for low token use per task.
Available in EULanguage Models
Release date n/a