GPT Sol
An OpenAI GPT model tier used for coding, code review and agentic workflows.
Commenters use it as a cost-effective daily driver for coding, code review, and planning or orchestration in multi-model setups,12345678 and praise how closely it follows system prompts and AGENTS.md instructions.910 The most common complaint is confident mistakes and hallucinations, with some saying newer versions regressed,111213 and a few find its style flat or lacking taste.1415
Why people recommend it
Cost value · 3
Often called the best bang for the buck and cheaper than comparable models.
“Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them”
“sol 6.1, best bang for buck.”
Show 1 more
“6.1-Sol is also cheaper than Sonnet 5.5 and much smarter.”
Code review · 3
Trusted for coding, review loops and plan execution.
“Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.”
“Using sol 6.1 for review loops at the end.”
Show 1 more
Orchestration · 3
Commonly assigned as planner, orchestrator or coordinator over cheaper models.
“I use gpt-6.1-sol high for planning/orchestration”
“Sol high as the orchestrator, Sol max as Planner & Reviewer”
Show 1 more
“Sol xhigh - for orchestration and implementation in complicated cases.”
Speed · 3
Described as fast and efficient in time and token usage.
“GPT Sol 6.1 is however, for me the most efficient model, both in terms of time to finish and token usage.”
“gpt Sol on medium or low effort to also be faster than opus”
Show 1 more
“6.1Sol High or X-High to build and even on huge projects with 10s of 1000s of pages it is fast and competent”
Instruction following · 2
Follows system prompts, tools, skills and AGENTS.md closely.
“I would much rather delegate it to GPT (more likely to draw on the harness for context, skills, instructions, and tools -> more consistency).”
“Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md”
Clear output · 3
Seen as direct, less verbose and easier to understand than Claude models.
“I do appreciate how more straight to the point Sol 5.6 seems to be”
“the gpt model series is far and above easier for me to understand the outputs of than claude”
Show 1 more
“Sol is generally preferred; it’s faster, cheaper, and less verbose.”
Caveats and dissent
Hallucinations · 3
Users report confident errors, hallucinations and regressions in newer versions.
“I'm deliberately avoiding Sol 6 right now due to some issues with hallucinations”
“This is astonishingly bad and it is nowhere close to what Sol 5.6 was a month ago.”
Show 1 more
“it consistently, confidently failed”
Task-specific failures · 3
Weak results reported on UI building, schematic editing and some research questions.
“I was using Sol originally which was ok but also irritating because it would mess up and make a buggy ui.”
“Those models were not capable of even schematic editing or building.”
Show 1 more
“I had asked an openai model (probably sol?) and got much worse answers”
Flat style · 2
Some find its communication stale or lacking taste.
“OAI models simplify have no taste.”
Usage limits · 1
One user burned a week of subscription usage in a day.
“I burned through my whole week in 1 day on Sol.”
Rabbit holing · 1
Can fixate on tiny implementation errors.
“Whereas Sol and Astra kept rabbit holing and looking for tiny implementation errors.”
Prompt blocking · 1
Blocked prompts during security hardening work.
“Sol/Opus keep snitching on me and blocking my prompts.”
Often compared with
5 moved from Claude Opus to GPT Sol
2 moved from GPT Sol to Claude Opus
5 moved from Claude Fable to GPT Sol
1 moved from GPT Sol to Claude Fable
2 moved from GPT Sol to GPT Luna
2 moved from GPT Astra to GPT Sol
2 moved from GPT Sol to GPT Astra
8 moved from Claude to GPT Sol
1 moved from GPT Sol to GPT Terra
1 moved from GPT Sol to DeepSeek V4 Flash
Who recommends it in their own words
Threads where it came up
- Ask HN: What default model do you use and why? 12 mentions
- GPT-6 Sol and Luna 9 mentions
- GPT-5.6 Sol Pricing Cut by 50% 8 mentions
- GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price 6 mentions
- DeepSeek-v4.1-Flash-1017 5 mentions
- GPT-6 Astra 5 mentions
Also written as GPT-5.6 Sol, GPT 5.6 Sol, Sol 5.6, GPT-6 Sol. Data: gpt-sol.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 45
“I would much rather delegate it to GPT (more likely to draw on the harness for context, skills, instructions, and tools -> more consistency).”
“has not been correctly fixed by sol-6.1 (medium)”
“lamenting how much worse Opus felt to use than Sol”
“I used Sol and GLM 5.3 Flash to produce a prototype of a tower defense game for myself, and it works just fine.”
“how stale gpt-sol is”
“Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them”
“6.1Sol High or X-High to build and even on huge projects with 10s of 1000s of pages it is fast and competent”
“Sol 6.1 is the big one for jobs that need significantly above average intelligence”
“I was able to transcribe the textbook and make straightforward corrections with Codex + GPT 6.1 Sol.”
“sol 6.1, best bang for buck.”
“Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md”
“Luna seems more willing than Sol”
“We used Wiz Atlas, Gemini 4 Argon and GPT-5.6 Sol throughout this research.”
“for webdev, deep research, code review, and a few other tasks I find the current 6.1 sol better”
“GPT Sol 6.1 is however, for me the most efficient model, both in terms of time to finish and token usage.”
“gpt Sol on medium or low effort to also be faster than opus”
“I use Sol on medium as the work I'm doing now needs more memory than it needs intelligence and it's working great.”
“it should elevate to a higher level model (Sol), and ask its review”
“I was trying out using Sol and liking the result”
“Using sol 6.1 for review loops at the end.”
“Most of my work is with GPT-5.6-Sol.”
“It also doesn't come up with as good solutions as Opus/Sol.”
“I use gpt-6.1-sol high for planning/orchestration”
“its about 10€ for sol to plan or debug and 5€ for luna to implement.”
“Presently I'm using sol-high for the default agent which does orchestration”
“now I'm on GPT-6 Sol. It's been a very good experience.”
“I'm no longer hitting any of the capability limits I used to experience”
“I used Sol 5.6 Extra High”
“I'm deliberately avoiding Sol 6 right now due to some issues with hallucinations”
“I typically ask GPT 6.1 Sol to review requirements”
“6.1-Sol is also cheaper than Sonnet 5.5 and much smarter.”
“Sol for real problems”
“OAI models simplify have no taste.”
“model = "gpt-6.1-sol"”
“I was using Sol originally which was ok but also irritating because it would mess up and make a buggy ui.”
“it's solid and cheaper”
“I expect this will be my new daily driver.”
“I've been using it exclusively since release and it feels like a cheaper astra to me.”
“Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt”
“I burned through my whole week in 1 day on Sol.”
“This is astonishingly bad and it is nowhere close to what Sol 5.6 was a month ago.”
“Sol high as the orchestrator, Sol max as Planner & Reviewer”
“gpt-5.6-sol high has been my go-to for months.”
“it consistently, confidently failed”
“Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.”