{"slug": "gpt-sol", "name": "GPT Sol", "category": "ai", "category_name": "AI models and assistants", "description": "An OpenAI GPT model tier used for coding, code review and agentic workflows.", "recommenders": 109, "commenters": 191, "warners": 24, "score": 92.922, "rising": 86.44, "first": "2026-03-30", "latest": "2026-10-31", "why": "Commenters use it as a cost-effective daily driver for coding, code review, and planning or orchestration in multi-model setups, and praise how closely it follows system prompts and AGENTS.md instructions.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 113, "neutral": 51, "compare": 20, "warn": 21}, "spark": [0, 0, 0, 0, 1, 0, 0, 0, 17, 36, 41, 18], "listed": true, "loved": ["Cost value", "Code review"], "caveat": "Hallucinations", "short_quote": "I just saw an amazing open-source alternative to Raycast[0] and just replaced it the other day.", "icon": true, "domain": null, "github_repo": "github.com/ospfranco/sol", "official_url": "https://github.com/ospfranco/sol", "pricing": "paid", "aliases": ["Sol", "GPT-5.6 Sol", "GPT 5.6 Sol", "Sol 5.6", "GPT-6 Sol", "GPT Sol", "Sol 6.1", "GPT 6.1 Sol", "GPT Sol 5.6", "GPT-5.6-Sol", "GPT-6.1 Sol", "Sol 6", "5.6 Sol", "GPT-Sol 6.1", "GPT Sol 6.1", "gpt-6.1-sol", "5.6 sol", "GPT 6 Sol", "GPT-5.6 SOL", "GPT 5.6-Sol", "sol-6.1", "Sol-6.1", "gpt-sol", "GPT6 Sol"], "verdict": [{"text": "Commenters use it as a cost-effective daily driver for coding, code review, and planning or orchestration in multi-model setups,", "cites": [{"id": 804031329, "item_id": 50251958, "url": "https://news.ycombinator.com/item?id=50251958", "by": "rockdoe", "date": "2026-10-26", "stance": "recommend", "quote": "Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them", "excerpt": "Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them, so I'd say really well so far. There's very few stuff that needs Astra/Opus level intelligence. Luna 6.1 would be amazing.", "context": "", "use_case": "cost-effective workflows", "switched_from": "", "story": "GPT-6.1 Terra"}, {"id": 798397648, "item_id": 49899853, "url": "https://news.ycombinator.com/item?id=49899853", "by": "jorblumesea", "date": "2026-09-29", "stance": "recommend", "quote": "it's solid and cheaper", "excerpt": "yeah I use sol constantly and have done maybe $15 of spend in the past week. it's solid and cheaper. this is at least 4-5 investigations, prs, whatever per day.", "context": "", "use_case": "investigations and PRs", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798569552, "item_id": 49910597, "url": "https://news.ycombinator.com/item?id=49910597", "by": "gertlabs", "date": "2026-09-30", "stance": "recommend", "quote": "6.1-Sol is also cheaper than Sonnet 5.5 and much smarter.", "excerpt": "We tested it across 100 unsaturated coding and engineering environments. Both Astra and 6.1-Sol are pretty comfortably ahead of Opus 5.5 in these types of evaluations, and both end up being cheaper than Opus via API usage. 6.1-Sol is also cheaper than Sonnet 5.5 and much smarter. The only verifiable domain Anthropic seems to be clearly ahead is chemistry (and perhaps also some unverifiable domains like being pleasant", "context": "", "use_case": "coding and engineering tasks", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 795268496, "item_id": 49704281, "url": "https://news.ycombinator.com/item?id=49704281", "by": "glimshe", "date": "2026-09-14", "stance": "recommend", "quote": "Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.", "excerpt": "I select my review model based on change complexity. If I'm confident that the change is localized (given that I always maintain an up to date and complete mental model of the application), I'll use Luna/Terra. For deeper changes, I use Sol/Astra. Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}, {"id": 802762896, "item_id": 50172681, "url": "https://news.ycombinator.com/item?id=50172681", "by": "singingtoday", "date": "2026-10-20", "stance": "recommend", "quote": "for webdev, deep research, code review, and a few other tasks I find the current 6.1 sol better", "excerpt": "I have no data to support this, but when working on SaaS systems I find anthropic models more effective. But for webdev, deep research, code review, and a few other tasks I find the current 6.1 sol better. But I use both. And I use codex more.", "context": "", "use_case": "webdev, research, code review", "switched_from": "", "story": "SlopCodeBench 10/2026 \u2013 Opus 5.5, Astra wins overall, xhigh considered harmful"}, {"id": 800197698, "item_id": 50012356, "url": "https://news.ycombinator.com/item?id=50012356", "by": "apitman", "date": "2026-10-08", "stance": "recommend", "quote": "I use gpt-6.1-sol high for planning/orchestration", "excerpt": "> With my OpenCode Go sub of $10/month, DeepSeek is basically unlimited My OpenCode Go monthly window was scheduled to reset this morning. It was sitting at 22% used despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks (I use gpt-6.1-sol high for planning/orchestration). I had 1.5 hours left so I fired up first 10, then 20, and finally 50 concurrent subagents all working on re", "context": "", "use_case": "planning and orchestration", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 796936434, "item_id": 49808527, "url": "https://news.ycombinator.com/item?id=49808527", "by": "jeremyjh", "date": "2026-09-22", "stance": "recommend", "quote": "Sol high as the orchestrator, Sol max as Planner & Reviewer", "excerpt": "You can get a lot of Codex usage out of that same sub on top of ChatGPT usage. Its a really good value and you can use that sub in any harness. In OMP I have Sol high as the orchestrator, Sol max as Planner & Reviewer, Luna max as task/coder. Very good setup. I'm on pro now and there are weekends when I use half a week's usage but I'll have 5 or 6 sessions going at once for many hours each day.", "context": "", "use_case": "orchestration, planning, review", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 796902128, "item_id": 49806383, "url": "https://news.ycombinator.com/item?id=49806383", "by": "apitman", "date": "2026-09-22", "stance": "recommend", "quote": "gpt-5.6-sol high has been my go-to for months.", "excerpt": "Similar for me. gpt-5.6-sol high has been my go-to for months. One of the reasons I'm pushing myself to try open models more is because it lends some level of guarantee I can continue to use the same tool as long as I want to. And I think we may just be getting to the point the open models are >= 5.6 Sol for coding.", "context": "", "use_case": "coding", "switched_from": "", "story": "GPT-6 Sol and Luna"}]}, {"text": " and praise how closely it follows system prompts and AGENTS.md instructions.", "cites": [{"id": 804993472, "item_id": 50312092, "url": "https://news.ycombinator.com/item?id=50312092", "by": "CharlieDigital", "date": "2026-10-31", "stance": "recommend", "quote": "I would much rather delegate it to GPT (more likely to draw on the harness for context, skills, instructions, and tools -> more consistency).", "excerpt": "I genuinely prefer OpenAI models for coding and building in general. My observation: GPT models are generally better at following the system prompt, tool use, and skill use. I use Paseo and there is a noticeable difference when GPT-6.1 Sol is driving the orchestrator vs Opus/Sonnet 5.5 driving the orchestrator. Anthropic models ignore the shared system prompt to use the Paseo tools for sub-agents and prefer to use it", "context": "", "use_case": "orchestrating coding agents", "switched_from": "", "story": "Anthropic to receive high 'junk' rating from credit agencies following IPO"}, {"id": 803487104, "item_id": 50217944, "url": "https://news.ycombinator.com/item?id=50217944", "by": "RussianCow", "date": "2026-10-24", "stance": "recommend", "quote": "Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md", "excerpt": "This is my experience as well. Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md, almost to an annoying point. (But I'd rather have that problem than the inverse.)", "context": "", "use_case": "following agent instructions", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}]}, {"text": " The most common complaint is confident mistakes and hallucinations, with some saying newer versions regressed,", "cites": [{"id": 799092849, "item_id": 49943303, "url": "https://news.ycombinator.com/item?id=49943303", "by": "Gareth321", "date": "2026-10-03", "stance": "warn", "quote": "I'm deliberately avoiding Sol 6 right now due to some issues with hallucinations", "excerpt": "Elvis wasn't the one calling for censorship. It was religious viewers - the audience. I'm not denying that there were examples of creatives - like punks - who called for censorship. I'm describing a massive shift in scale and scope. I did as you suggested because it's important to have a broad and balanced view of the issues. I used the query \"Do you believe the scale and scope of creatives calling for censorship is ", "context": "", "use_case": "", "switched_from": "", "story": "Newgrounds.com \u2013 A community of games, music, and art"}, {"id": 798177712, "item_id": 49886107, "url": "https://news.ycombinator.com/item?id=49886107", "by": "cromka", "date": "2026-09-28", "stance": "warn", "quote": "This is astonishingly bad and it is nowhere close to what Sol 5.6 was a month ago.", "excerpt": "Not even GPT6 Sol cannot match DeepSeek 4.1 in my work, with outrageous bugs. Don't get me started on Luna. From my today's session with Sol: - it actually failed to correctly understand a simple English grammar and logical implication of it, then when challenged it admitted its mistake but couldn't explain why it made it. - for the code I am working on, I asked to create two PRs for the two small features (couple li", "context": "", "use_case": "coding", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 795344272, "item_id": 49709017, "url": "https://news.ycombinator.com/item?id=49709017", "by": "trencedamp", "date": "2026-09-15", "stance": "warn", "quote": "it consistently, confidently failed", "excerpt": "It's not the word I'm taking issue with, it's the concept. From your post you are surprised that the model makes mistakes and does not recognize things, and these are only surprising if you imagine the models to be thinking about things. We can argue about words like thinking and intelligence all day long, and so can an LLM. But at the end of the day the LLM is only mimicking the processes you or I use. Last week I t", "context": "", "use_case": "counting letters in a long string", "switched_from": "", "story": "Pion, an agent designed to run any company autonomously"}]}, {"text": " and a few find its style flat or lacking taste.", "cites": [{"id": 804323313, "item_id": 50270207, "url": "https://news.ycombinator.com/item?id=50270207", "by": "wyre", "date": "2026-10-28", "stance": "warn", "quote": "how stale gpt-sol is", "excerpt": "I used MiMo for a couple of days waiting for my Codex sub to reset. I really liked it, its communication style was a breath of fresh air compared to how stale gpt-sol is.", "context": "", "use_case": "", "switched_from": "", "story": "Artificial Analysis Intelligence Index v5"}, {"id": 798483346, "item_id": 49905209, "url": "https://news.ycombinator.com/item?id=49905209", "by": "laurels-marts", "date": "2026-09-30", "stance": "warn", "quote": "OAI models simplify have no taste.", "excerpt": "100% in agreement. I pay for OAI sub and also use Codex exclusively at work for the past 8 months. I tried out fable 5.1 the day it was released and coming from gpt-5.6-sol I was truly mind blown (both in terms of code and prose it was generating - outputs I could finally enjoy reading and looking at). Then when opus 5.5 came out, again same thing + far cheaper and faster. I went from using OAI exclusively the entire", "context": "", "use_case": "coding", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}]}], "praise": [{"label": "Cost value", "summary": "Often called the best bang for the buck and cheaper than comparable models.", "quotes": [{"id": 804031329, "item_id": 50251958, "url": "https://news.ycombinator.com/item?id=50251958", "by": "rockdoe", "date": "2026-10-26", "stance": "recommend", "quote": "Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them", "excerpt": "Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them, so I'd say really well so far. There's very few stuff that needs Astra/Opus level intelligence. Luna 6.1 would be amazing.", "context": "", "use_case": "cost-effective workflows", "switched_from": "", "story": "GPT-6.1 Terra"}, {"id": 803556578, "item_id": 50222286, "url": "https://news.ycombinator.com/item?id=50222286", "by": "potsandpans", "date": "2026-10-24", "stance": "recommend", "quote": "sol 6.1, best bang for buck.", "excerpt": "I recently just crossed the threshold for 100% decompilation of a PlayStation game. I used a lot of models in the process, and I feel like this is a good demonstration the \"sauce\" that opus has. The harness I've built for it is pretty straight forward: split binary into individual functions, give the agent a bunch of small tools to diff output with underlying binary, make them wrote clean c until there's a certified ", "context": "decompiling a PlayStation game", "use_case": "decompiling functions", "switched_from": "", "story": "Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line"}, {"id": 798569552, "item_id": 49910597, "url": "https://news.ycombinator.com/item?id=49910597", "by": "gertlabs", "date": "2026-09-30", "stance": "recommend", "quote": "6.1-Sol is also cheaper than Sonnet 5.5 and much smarter.", "excerpt": "We tested it across 100 unsaturated coding and engineering environments. Both Astra and 6.1-Sol are pretty comfortably ahead of Opus 5.5 in these types of evaluations, and both end up being cheaper than Opus via API usage. 6.1-Sol is also cheaper than Sonnet 5.5 and much smarter. The only verifiable domain Anthropic seems to be clearly ahead is chemistry (and perhaps also some unverifiable domains like being pleasant", "context": "", "use_case": "coding and engineering tasks", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}]}, {"label": "Code review", "summary": "Trusted for coding, review loops and plan execution.", "quotes": [{"id": 795268496, "item_id": 49704281, "url": "https://news.ycombinator.com/item?id=49704281", "by": "glimshe", "date": "2026-09-14", "stance": "recommend", "quote": "Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.", "excerpt": "I select my review model based on change complexity. If I'm confident that the change is localized (given that I always maintain an up to date and complete mental model of the application), I'll use Luna/Terra. For deeper changes, I use Sol/Astra. Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}, {"id": 802068867, "item_id": 50129304, "url": "https://news.ycombinator.com/item?id=50129304", "by": "egeozcan", "date": "2026-10-17", "stance": "recommend", "quote": "Using sol 6.1 for review loops at the end.", "excerpt": "I use Opus for planning, through API on medium thinking, it doesn't cost anything significant and the rest can be happily done with deepseek or glm flash. Using sol 6.1 for review loops at the end. All running on pi. A very complex task tends to cost like 10$ max. It's comparable to a subscription with enormous flexibility. It's a steal.", "context": "", "use_case": "code review", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 796994081, "item_id": 49812130, "url": "https://news.ycombinator.com/item?id=49812130", "by": "slopinthebag", "date": "2026-09-23", "stance": "recommend", "quote": "sol was much closer to reality", "excerpt": "shocking. the code it generated, while technically working, was entirely garbage. i used it for code review and it flagged twenty issues, sol checked the review and found 75% of them were hallucinations. sol was much closer to reality. i no longer trust benchmarks at all because of it.", "context": "", "use_case": "checking code review", "switched_from": "", "story": "GPT-6 Sol and Luna"}]}, {"label": "Orchestration", "summary": "Commonly assigned as planner, orchestrator or coordinator over cheaper models.", "quotes": [{"id": 800197698, "item_id": 50012356, "url": "https://news.ycombinator.com/item?id=50012356", "by": "apitman", "date": "2026-10-08", "stance": "recommend", "quote": "I use gpt-6.1-sol high for planning/orchestration", "excerpt": "> With my OpenCode Go sub of $10/month, DeepSeek is basically unlimited My OpenCode Go monthly window was scheduled to reset this morning. It was sitting at 22% used despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks (I use gpt-6.1-sol high for planning/orchestration). I had 1.5 hours left so I fired up first 10, then 20, and finally 50 concurrent subagents all working on re", "context": "", "use_case": "planning and orchestration", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 796936434, "item_id": 49808527, "url": "https://news.ycombinator.com/item?id=49808527", "by": "jeremyjh", "date": "2026-09-22", "stance": "recommend", "quote": "Sol high as the orchestrator, Sol max as Planner & Reviewer", "excerpt": "You can get a lot of Codex usage out of that same sub on top of ChatGPT usage. Its a really good value and you can use that sub in any harness. In OMP I have Sol high as the orchestrator, Sol max as Planner & Reviewer, Luna max as task/coder. Very good setup. I'm on pro now and there are weekends when I use half a week's usage but I'll have 5 or 6 sessions going at once for many hours each day.", "context": "", "use_case": "orchestration, planning, review", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 795549793, "item_id": 49721862, "url": "https://news.ycombinator.com/item?id=49721862", "by": "pod_krad", "date": "2026-09-16", "stance": "recommend", "quote": "Sol xhigh - for orchestration and implementation in complicated cases.", "excerpt": "Astra xhigh - for specs, analysis. Sol xhigh - for orchestration and implementation in complicated cases. Terra, Luna - for implementation in simple routine cases.", "context": "", "use_case": "orchestration and complex implementation", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"label": "Speed", "summary": "Described as fast and efficient in time and token usage.", "quotes": [{"id": 802516450, "item_id": 50157278, "url": "https://news.ycombinator.com/item?id=50157278", "by": "sinuhe69", "date": "2026-10-19", "stance": "recommend", "quote": "GPT Sol 6.1 is however, for me the most efficient model, both in terms of time to finish and token usage.", "excerpt": "On Openrouter at least, all Mino models are painfully slow. TTFT is many seconds and the models spend a lot of time on its CoT, making effective output per second many times slower than DS Flash. On Flash 4.1, I constantly get ~180 tok/s and the real work got done fast, too. So even when the Mimo 2.6 are smart models, I use them much less recently. Not to mention their costs are also higher than DS. For me, the DeepS", "context": "", "use_case": "", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802192640, "item_id": 50137040, "url": "https://news.ycombinator.com/item?id=50137040", "by": "markasoftware", "date": "2026-10-18", "stance": "recommend", "quote": "gpt Sol on medium or low effort to also be faster than opus", "excerpt": "Fwiw I find gpt Sol on medium or low effort to also be faster than opus to the point of helping me not check out, but is quite smarter than dsv4 flash", "context": "", "use_case": "coding", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 804004273, "item_id": 50250267, "url": "https://news.ycombinator.com/item?id=50250267", "by": "r0fl", "date": "2026-10-26", "stance": "recommend", "quote": "6.1Sol High or X-High to build and even on huge projects with 10s of 1000s of pages it is fast and competent", "excerpt": "I cannot figure it out either. I use Astra Ultra to plan and 6.1Sol High or X-High to build and even on huge projects with 10s of 1000s of pages it is fast and competent", "context": "", "use_case": "building code", "switched_from": "", "story": "GPT-6.1 Terra"}]}, {"label": "Instruction following", "summary": "Follows system prompts, tools, skills and AGENTS.md closely.", "quotes": [{"id": 804993472, "item_id": 50312092, "url": "https://news.ycombinator.com/item?id=50312092", "by": "CharlieDigital", "date": "2026-10-31", "stance": "recommend", "quote": "I would much rather delegate it to GPT (more likely to draw on the harness for context, skills, instructions, and tools -> more consistency).", "excerpt": "I genuinely prefer OpenAI models for coding and building in general. My observation: GPT models are generally better at following the system prompt, tool use, and skill use. I use Paseo and there is a noticeable difference when GPT-6.1 Sol is driving the orchestrator vs Opus/Sonnet 5.5 driving the orchestrator. Anthropic models ignore the shared system prompt to use the Paseo tools for sub-agents and prefer to use it", "context": "", "use_case": "orchestrating coding agents", "switched_from": "", "story": "Anthropic to receive high 'junk' rating from credit agencies following IPO"}, {"id": 803487104, "item_id": 50217944, "url": "https://news.ycombinator.com/item?id=50217944", "by": "RussianCow", "date": "2026-10-24", "stance": "recommend", "quote": "Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md", "excerpt": "This is my experience as well. Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md, almost to an annoying point. (But I'd rather have that problem than the inverse.)", "context": "", "use_case": "following agent instructions", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}]}, {"label": "Clear output", "summary": "Seen as direct, less verbose and easier to understand than Claude models.", "quotes": [{"id": 794779520, "item_id": 49673720, "url": "https://news.ycombinator.com/item?id=49673720", "by": "r_lee", "date": "2026-09-12", "stance": "recommend", "quote": "I do appreciate how more straight to the point Sol 5.6 seems to be", "excerpt": "I recently switched over to GPT from Claude once my subscription lapsed, it's too early to tell, but I do appreciate how more straight to the point Sol 5.6 seems to be, vs. the insane wordslop machine Opus tends to be. It's really strange because when Opus 5 released, there were some that pointed this out, but a bunch simply said it was the best and as good as Fable etc etc. but for me, it caused me to get very demot", "context": "", "use_case": "coding", "switched_from": "Claude", "story": "Ask HN: What default model do you use and why?"}, {"id": 797575968, "item_id": 49848498, "url": "https://news.ycombinator.com/item?id=49848498", "by": "trevorkiw", "date": "2026-09-25", "stance": "recommend", "quote": "the gpt model series is far and above easier for me to understand the outputs of than claude", "excerpt": "gpt 5.6 sol. assuming the prose is similar with the gpt 6 series, if it was available at work that's probably what i'd be using instead. i'm well aware of opus and fable's superior engineering qualities for writing code, however i find that as a swe who needs to understand every engineering decision that is made and is held accountable for their actions for safety critical code, understanding trumps speed every time.", "context": "SWE on safety critical code", "use_case": "safety-critical software engineering", "switched_from": "", "story": "Ask HN: Which model do you use for work?"}, {"id": 794782802, "item_id": 49673925, "url": "https://news.ycombinator.com/item?id=49673925", "by": "verisimilidude", "date": "2026-09-12", "stance": "recommend", "quote": "Sol is generally preferred; it\u2019s faster, cheaper, and less verbose.", "excerpt": "It really depends. For high-level strategic conversations, I use Fable. For planning, I use Opus or Sol. Sol is generally preferred; it\u2019s faster, cheaper, and less verbose. But I still find Opus more capable on the most nuanced or complex tasks. For planned implementation, I use Sonnet. For one-shot unplanned implementation, I\u2019ll use whatever model seems best for the task. FWIW, I\u2019m increasingly turning to Grok here.", "context": "", "use_case": "planning", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}], "caveats": [{"label": "Hallucinations", "summary": "Users report confident errors, hallucinations and regressions in newer versions.", "quotes": [{"id": 799092849, "item_id": 49943303, "url": "https://news.ycombinator.com/item?id=49943303", "by": "Gareth321", "date": "2026-10-03", "stance": "warn", "quote": "I'm deliberately avoiding Sol 6 right now due to some issues with hallucinations", "excerpt": "Elvis wasn't the one calling for censorship. It was religious viewers - the audience. I'm not denying that there were examples of creatives - like punks - who called for censorship. I'm describing a massive shift in scale and scope. I did as you suggested because it's important to have a broad and balanced view of the issues. I used the query \"Do you believe the scale and scope of creatives calling for censorship is ", "context": "", "use_case": "", "switched_from": "", "story": "Newgrounds.com \u2013 A community of games, music, and art"}, {"id": 798177712, "item_id": 49886107, "url": "https://news.ycombinator.com/item?id=49886107", "by": "cromka", "date": "2026-09-28", "stance": "warn", "quote": "This is astonishingly bad and it is nowhere close to what Sol 5.6 was a month ago.", "excerpt": "Not even GPT6 Sol cannot match DeepSeek 4.1 in my work, with outrageous bugs. Don't get me started on Luna. From my today's session with Sol: - it actually failed to correctly understand a simple English grammar and logical implication of it, then when challenged it admitted its mistake but couldn't explain why it made it. - for the code I am working on, I asked to create two PRs for the two small features (couple li", "context": "", "use_case": "coding", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 795344272, "item_id": 49709017, "url": "https://news.ycombinator.com/item?id=49709017", "by": "trencedamp", "date": "2026-09-15", "stance": "warn", "quote": "it consistently, confidently failed", "excerpt": "It's not the word I'm taking issue with, it's the concept. From your post you are surprised that the model makes mistakes and does not recognize things, and these are only surprising if you imagine the models to be thinking about things. We can argue about words like thinking and intelligence all day long, and so can an LLM. But at the end of the day the LLM is only mimicking the processes you or I use. Last week I t", "context": "", "use_case": "counting letters in a long string", "switched_from": "", "story": "Pion, an agent designed to run any company autonomously"}]}, {"label": "Task-specific failures", "summary": "Weak results reported on UI building, schematic editing and some research questions.", "quotes": [{"id": 798439075, "item_id": 49902442, "url": "https://news.ycombinator.com/item?id=49902442", "by": "srmckee", "date": "2026-09-29", "stance": "warn", "quote": "I was using Sol originally which was ok but also irritating because it would mess up and make a buggy ui.", "excerpt": "Hey HN! Spencer here - wanted to share a token compression tool that I\u2019ve built for myself to save 30% costs on codex! I've been building a ton of web apps to replace saas i'm too cheap to keep paying for like chatgpt, notion, myfitnesspal, etc. and so i can get the exact features i want (+ no ads!). I was using Sol originally which was ok but also irritating because it would mess up and make a buggy ui. so I switche", "context": "", "use_case": "coding model", "switched_from": "", "story": "Show HN: Free token compression CLI, saves codex bills by 30%"}, {"id": 795840130, "item_id": 49740008, "url": "https://news.ycombinator.com/item?id=49740008", "by": "the__alchemist", "date": "2026-09-17", "stance": "warn", "quote": "Those models were not capable of even schematic editing or building.", "excerpt": "Hey - I ordered a board yesterday. I used Opus 5, Astra, and Sol at points to QC the board, and offer power routing advice. (The MCU is a brand new one, and poorly documented unless they send you an NDA and the full docs). Those models were not capable of even schematic editing or building. (I didn't bother with routing or placement). They made gross errors all over the schematic; changed the known-working parts and ", "context": "", "use_case": "PCB schematic QC", "switched_from": "", "story": "This PCB is brought to you by Fable 5"}, {"id": 796764785, "item_id": 49797799, "url": "https://news.ycombinator.com/item?id=49797799", "by": "StevenWaterman", "date": "2026-09-22", "stance": "warn", "quote": "I had asked an openai model (probably sol?) and got much worse answers", "excerpt": "Cool I stand corrected, I had asked an openai model (probably sol?) and got much worse answers, broadly: - Listing the reasons why Hollywood claimed to want to law, like not giving premature ideas that a movie would flop, without pointing out how much Hollywood benefits from the information symmetry - Saying yes the gratuities affect pay but not as straightforwardly as you would expect, quoting the \"all gratuities go", "context": "", "use_case": "fact-checking questions", "switched_from": "", "story": "Nearly half of young people trust AI over a human for fact-checking, report says"}]}, {"label": "Flat style", "summary": "Some find its communication stale or lacking taste.", "quotes": [{"id": 804323313, "item_id": 50270207, "url": "https://news.ycombinator.com/item?id=50270207", "by": "wyre", "date": "2026-10-28", "stance": "warn", "quote": "how stale gpt-sol is", "excerpt": "I used MiMo for a couple of days waiting for my Codex sub to reset. I really liked it, its communication style was a breath of fresh air compared to how stale gpt-sol is.", "context": "", "use_case": "", "switched_from": "", "story": "Artificial Analysis Intelligence Index v5"}, {"id": 798483346, "item_id": 49905209, "url": "https://news.ycombinator.com/item?id=49905209", "by": "laurels-marts", "date": "2026-09-30", "stance": "warn", "quote": "OAI models simplify have no taste.", "excerpt": "100% in agreement. I pay for OAI sub and also use Codex exclusively at work for the past 8 months. I tried out fable 5.1 the day it was released and coming from gpt-5.6-sol I was truly mind blown (both in terms of code and prose it was generating - outputs I could finally enjoy reading and looking at). Then when opus 5.5 came out, again same thing + far cheaper and faster. I went from using OAI exclusively the entire", "context": "", "use_case": "coding", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}]}, {"label": "Usage limits", "summary": "One user burned a week of subscription usage in a day.", "quotes": [{"id": 798274050, "item_id": 49892128, "url": "https://news.ycombinator.com/item?id=49892128", "by": "joshstrange", "date": "2026-09-29", "stance": "warn", "quote": "I burned through my whole week in 1 day on Sol.", "excerpt": "I can believe that (20x is not 4x of the 5x plan) but I\u2019ve been on the $100/mo Codex and the $200/mo Claude plans and I can run the whole week with multiple agents on Claude, mean while I burned through my whole week in 1 day on Sol. Yes, it\u2019s half the price of my Claude plan but even if OpenAi actually gave 4x usage for the next plan up (which they just halved), it would still be less than the usage I get from Claud", "context": "", "use_case": "coding", "switched_from": "", "story": "OpenAI: Tomorrow we are re-opening the Pro $200 subscription"}]}, {"label": "Rabbit holing", "summary": "Can fixate on tiny implementation errors.", "quotes": [{"id": 796933585, "item_id": 49808349, "url": "https://news.ycombinator.com/item?id=49808349", "by": "spockz", "date": "2026-09-22", "stance": "warn", "quote": "Whereas Sol and Astra kept rabbit holing and looking for tiny implementation errors.", "excerpt": "I use it for golang, and it is fantastic. Incredibly fast. It seems the llm and I \u201cunderstand\u201d each other. I have to be less careful in my exact phrasing. It kind of just does what I want and expect. When I ask for an explanation it adds the right amount of detail. Of course, some of the material is new to me so subtle errors are hard to spot. But at least I\u2019ve caught Terra and Sol on inconsistent messaging. Also I\u2019v", "context": "", "use_case": "", "switched_from": "", "story": "GPT-6 Sol and Luna"}]}, {"label": "Prompt blocking", "summary": "Blocked prompts during security hardening work.", "quotes": [{"id": 794778899, "item_id": 49673681, "url": "https://news.ycombinator.com/item?id=49673681", "by": "conradludgate", "date": "2026-09-12", "stance": "warn", "quote": "Sol/Opus keep snitching on me and blocking my prompts.", "excerpt": "For personal use, Luna-xhigh on the Plus subscription. I only ever hit my limits if I decide to use Sol or Astra for a bit more quality. For work, given we pay API pricing anyway, I've been happy experimenting with K3-medium as my default, and now I'm planning on trying GLM 5.3 as well. Sol-medium is my fallback for my work purposes but I occasionally use Opus 4.8 as an additional reviewer. I like the open weight mod", "context": "does security hardening of own service", "use_case": "security hardening", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}], "who": [{"by": "elxr", "context": "uses models daily on a modest budget", "url": "https://news.ycombinator.com/item?id=50249354", "date": "2026-10-26"}, {"by": "potsandpans", "context": "decompiling a PlayStation game", "url": "https://news.ycombinator.com/item?id=50222286", "date": "2026-10-24"}, {"by": "zapnuk", "context": "uses API billing at work", "url": "https://news.ycombinator.com/item?id=50004455", "date": "2026-10-08"}, {"by": "chilmers", "context": "has subscriptions to both Anthropic and OpenAI", "url": "https://news.ycombinator.com/item?id=49976571", "date": "2026-10-06"}, {"by": "pettijohn", "context": "runs local models on Strix Halo 128GB", "url": "https://news.ycombinator.com/item?id=49929275", "date": "2026-10-02"}, {"by": "trevorkiw", "context": "SWE on safety critical code", "url": "https://news.ycombinator.com/item?id=49848498", "date": "2026-09-25"}, {"by": "schipperai", "context": "heavy Claude Code and Codex user", "url": "https://news.ycombinator.com/item?id=49809738", "date": "2026-09-22"}, {"by": "Gareth321", "context": "paying Codex subscriber", "url": "https://news.ycombinator.com/item?id=49798420", "date": "2026-09-22"}, {"by": "unsupp0rted", "context": "pays $200/mo", "url": "https://news.ycombinator.com/item?id=49798124", "date": "2026-09-22"}, {"by": "pkulak", "context": "$20 subscription user", "url": "https://news.ycombinator.com/item?id=49706632", "date": "2026-09-15"}, {"by": "jakevoytko", "context": "former Google Docs team member", "url": "https://news.ycombinator.com/item?id=49697215", "date": "2026-09-14"}, {"by": "m0rde", "context": "enterprise pay-as-you-go, cost conscious", "url": "https://news.ycombinator.com/item?id=49673682", "date": "2026-09-12"}], "tags": ["for coding agents", "for code review", "for orchestration", "for budget-conscious devs"], "use_cases": ["coding", "code review", "orchestrating coding agents", "planning", "harder tasks", "ai coding"], "compared": [{"slug": "claude-opus", "name": "Claude Opus", "co": 79, "switched_from": 5, "switched_to": 2}, {"slug": "claude-fable", "name": "Claude Fable", "co": 56, "switched_from": 5, "switched_to": 1}, {"slug": "gpt-luna", "name": "GPT Luna", "co": 51, "switched_from": 0, "switched_to": 2}, {"slug": "gpt-astra", "name": "GPT Astra", "co": 45, "switched_from": 2, "switched_to": 2}, {"slug": "claude", "name": "Claude", "co": 28, "switched_from": 8, "switched_to": 0}, {"slug": "codex", "name": "Codex", "co": 42, "switched_from": 0, "switched_to": 0}, {"slug": "gpt-terra", "name": "GPT Terra", "co": 20, "switched_from": 0, "switched_to": 1}, {"slug": "deepseek-v4-flash", "name": "DeepSeek V4 Flash", "co": 17, "switched_from": 0, "switched_to": 1}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49672966", "title": "Ask HN: What default model do you use and why?", "mentions": 12}, {"url": "https://news.ycombinator.com/item?id=49805509", "title": "GPT-6 Sol and Luna", "mentions": 9}, {"url": "https://news.ycombinator.com/item?id=49337602", "title": "GPT-5.6 Sol Pricing Cut by 50%", "mentions": 8}, {"url": "https://news.ycombinator.com/item?id=49896586", "title": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price", "mentions": 6}, {"url": "https://news.ycombinator.com/item?id=50129045", "title": "DeepSeek-v4.1-Flash-1017", "mentions": 5}, {"url": "https://news.ycombinator.com/item?id=49554643", "title": "GPT-6 Astra", "mentions": 5}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {"recommend": 1}, "2026-04": {}, "2026-05": {}, "2026-06": {}, "2026-07": {"recommend": 17, "compare": 3, "warn": 6, "neutral": 8}, "2026-08": {"neutral": 16, "recommend": 36, "warn": 4, "compare": 8}, "2026-09": {"neutral": 21, "recommend": 41, "compare": 6, "warn": 10}, "2026-10": {"neutral": 6, "recommend": 18, "compare": 3, "warn": 1}}, "quotes": [{"id": 804993472, "item_id": 50312092, "url": "https://news.ycombinator.com/item?id=50312092", "by": "CharlieDigital", "date": "2026-10-31", "stance": "recommend", "quote": "I would much rather delegate it to GPT (more likely to draw on the harness for context, skills, instructions, and tools -> more consistency).", "excerpt": "I genuinely prefer OpenAI models for coding and building in general. My observation: GPT models are generally better at following the system prompt, tool use, and skill use. I use Paseo and there is a noticeable difference when GPT-6.1 Sol is driving the orchestrator vs Opus/Sonnet 5.5 driving the orchestrator. Anthropic models ignore the shared system prompt to use the Paseo tools for sub-agents and prefer to use it", "context": "", "use_case": "orchestrating coding agents", "switched_from": "", "story": "Anthropic to receive high 'junk' rating from credit agencies following IPO"}, {"id": 804621122, "item_id": 50288820, "url": "https://news.ycombinator.com/item?id=50288820", "by": "gnufied", "date": "2026-10-29", "stance": "compare", "quote": "has not been correctly fixed by sol-6.1 (medium)", "excerpt": "hmm, I haven't tried Carbon yet. But I have been using Argon on and off via agent-shell - https://github.com/xenodium/agent-shell and unofficial ACP client. In my limited usage, it has fixed issues which has not been correctly fixed by sol-6.1 (medium) - https://github.com/auto-gnufied/markdown-modern/commit/a189a06f157954e97f45e6b1e9b64a2a68acfdcd I am also using it via Google pro tier and did not immediately run in", "context": "", "use_case": "", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}, {"id": 804479953, "item_id": 50279997, "url": "https://news.ycombinator.com/item?id=50279997", "by": "RestartKernel", "date": "2026-10-28", "stance": "compare", "quote": "lamenting how much worse Opus felt to use than Sol", "excerpt": "If the entire industry stopped training new models, and just served Opus 5.5, I would be perfectly satisfied. I don't know about its absolute performance, but Opus 5.5 is very \"tasteful\". It came out a week after I switched from Codex and started lamenting how much worse Opus felt to use than Sol, but whatever Anthropic did to fix it worked. Hopefully this isn't a fluke and they'll be able to keep on track with futur", "context": "", "use_case": "AI coding", "switched_from": "", "story": "Opus 5.5 is awesome"}, {"id": 804382224, "item_id": 50273889, "url": "https://news.ycombinator.com/item?id=50273889", "by": "Sphax", "date": "2026-10-28", "stance": "neutral", "quote": "I used Sol and GLM 5.3 Flash to produce a prototype of a tower defense game for myself, and it works just fine.", "excerpt": "I'm not looking forward to playing vibe coded games, but yes I feel like models are good enough today to write small scoped games. I used Sol and GLM 5.3 Flash to produce a prototype of a tower defense game for myself, and it works just fine. I _am_ looking forward however to game developers using AI to debug and optimize their games. Let's squeeze all the performance we can get from existing hardware.", "context": "", "use_case": "prototyping a game", "switched_from": "", "story": "The October inflection point for vibe game development"}, {"id": 804323313, "item_id": 50270207, "url": "https://news.ycombinator.com/item?id=50270207", "by": "wyre", "date": "2026-10-28", "stance": "warn", "quote": "how stale gpt-sol is", "excerpt": "I used MiMo for a couple of days waiting for my Codex sub to reset. I really liked it, its communication style was a breath of fresh air compared to how stale gpt-sol is.", "context": "", "use_case": "", "switched_from": "", "story": "Artificial Analysis Intelligence Index v5"}, {"id": 804031329, "item_id": 50251958, "url": "https://news.ycombinator.com/item?id=50251958", "by": "rockdoe", "date": "2026-10-26", "stance": "recommend", "quote": "Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them", "excerpt": "Luna and Sol-6.1 are crazy cost effective and all flows here have switched to them, so I'd say really well so far. There's very few stuff that needs Astra/Opus level intelligence. Luna 6.1 would be amazing.", "context": "", "use_case": "cost-effective workflows", "switched_from": "", "story": "GPT-6.1 Terra"}, {"id": 804004273, "item_id": 50250267, "url": "https://news.ycombinator.com/item?id=50250267", "by": "r0fl", "date": "2026-10-26", "stance": "recommend", "quote": "6.1Sol High or X-High to build and even on huge projects with 10s of 1000s of pages it is fast and competent", "excerpt": "I cannot figure it out either. I use Astra Ultra to plan and 6.1Sol High or X-High to build and even on huge projects with 10s of 1000s of pages it is fast and competent", "context": "", "use_case": "building code", "switched_from": "", "story": "GPT-6.1 Terra"}, {"id": 803989665, "item_id": 50249354, "url": "https://news.ycombinator.com/item?id=50249354", "by": "elxr", "date": "2026-10-26", "stance": "recommend", "quote": "Sol 6.1 is the big one for jobs that need significantly above average intelligence", "excerpt": "If you daily all these models (and have a modest budget, not some >$400 corporate allowance), then it's pretty obvious that Luna is small and fast, and Sol 6.1 is the big one for jobs that need significantly above average intelligence, comparatively to a human engineer. I use both constantly. Astra is really not relevant for the average dev except to do some moonshot task with a vague verification loop. Sure Sol 6.1 ", "context": "uses models daily on a modest budget", "use_case": "hard engineering tasks", "switched_from": "", "story": "GPT-6.1 Terra"}, {"id": 803909681, "item_id": 50244355, "url": "https://news.ycombinator.com/item?id=50244355", "by": "mathewjacob", "date": "2026-10-26", "stance": "recommend", "quote": "I was able to transcribe the textbook and make straightforward corrections with Codex + GPT 6.1 Sol.", "excerpt": "I've built what I believe is currently the best free, online resource for learning Malayalam from scratch. It is based on a textbook and related materials made available by the University of Texas at Austin (UT) which are available under the Creative Commons License. It is a full University course with grammar lessons, vocabulary, and audio clips. I've used those to also build Anki decks for flash cards. A significan", "context": "", "use_case": "textbook transcription and corrections", "switched_from": "", "story": "Malayalam: A University Course"}, {"id": 803556578, "item_id": 50222286, "url": "https://news.ycombinator.com/item?id=50222286", "by": "potsandpans", "date": "2026-10-24", "stance": "recommend", "quote": "sol 6.1, best bang for buck.", "excerpt": "I recently just crossed the threshold for 100% decompilation of a PlayStation game. I used a lot of models in the process, and I feel like this is a good demonstration the \"sauce\" that opus has. The harness I've built for it is pretty straight forward: split binary into individual functions, give the agent a bunch of small tools to diff output with underlying binary, make them wrote clean c until there's a certified ", "context": "decompiling a PlayStation game", "use_case": "decompiling functions", "switched_from": "", "story": "Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line"}, {"id": 803487104, "item_id": 50217944, "url": "https://news.ycombinator.com/item?id=50217944", "by": "RussianCow", "date": "2026-10-24", "stance": "recommend", "quote": "Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md", "excerpt": "This is my experience as well. Sol and Astra are by far the best of the models I use at following the instructions in AGENTS.md, almost to an annoying point. (But I'd rather have that problem than the inverse.)", "context": "", "use_case": "following agent instructions", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 803262145, "item_id": 50203884, "url": "https://news.ycombinator.com/item?id=50203884", "by": "dawnerd", "date": "2026-10-23", "stance": "compare", "quote": "Luna seems more willing than Sol", "excerpt": "Not very interesting, just being vague to not give it any context that it might be doing security research. Treating it like a game also works fairly well. It's hard to get more detailed about how since I use it for work and don't want it patched out, ya know? Splitting it up between turns, like Luna seems more willing than Sol. If you get blocked better to take what context you have and start a new chat and try to r", "context": "Uses it for work", "use_case": "security research prompting", "switched_from": "", "story": "Jailbreaking the Switch 2"}, {"id": 802900737, "item_id": 50181296, "url": "https://news.ycombinator.com/item?id=50181296", "by": "tchalla", "date": "2026-10-21", "stance": "neutral", "quote": "We used Wiz Atlas, Gemini 4 Argon and GPT-5.6 Sol throughout this research.", "excerpt": "> We used Wiz Atlas, Gemini 4 Argon and GPT-5.6 Sol throughout this research. Notably, Gemini 4 Argon made an invaluable contribution to finding the vulnerability, exploit development, attack detection, and validation of the patch.", "context": "", "use_case": "vulnerability research", "switched_from": "", "story": "PHP2Shell: AI uncovers critical RCE in PHP"}, {"id": 802762896, "item_id": 50172681, "url": "https://news.ycombinator.com/item?id=50172681", "by": "singingtoday", "date": "2026-10-20", "stance": "recommend", "quote": "for webdev, deep research, code review, and a few other tasks I find the current 6.1 sol better", "excerpt": "I have no data to support this, but when working on SaaS systems I find anthropic models more effective. But for webdev, deep research, code review, and a few other tasks I find the current 6.1 sol better. But I use both. And I use codex more.", "context": "", "use_case": "webdev, research, code review", "switched_from": "", "story": "SlopCodeBench 10/2026 \u2013 Opus 5.5, Astra wins overall, xhigh considered harmful"}, {"id": 802516450, "item_id": 50157278, "url": "https://news.ycombinator.com/item?id=50157278", "by": "sinuhe69", "date": "2026-10-19", "stance": "recommend", "quote": "GPT Sol 6.1 is however, for me the most efficient model, both in terms of time to finish and token usage.", "excerpt": "On Openrouter at least, all Mino models are painfully slow. TTFT is many seconds and the models spend a lot of time on its CoT, making effective output per second many times slower than DS Flash. On Flash 4.1, I constantly get ~180 tok/s and the real work got done fast, too. So even when the Mimo 2.6 are smart models, I use them much less recently. Not to mention their costs are also higher than DS. For me, the DeepS", "context": "", "use_case": "", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802192640, "item_id": 50137040, "url": "https://news.ycombinator.com/item?id=50137040", "by": "markasoftware", "date": "2026-10-18", "stance": "recommend", "quote": "gpt Sol on medium or low effort to also be faster than opus", "excerpt": "Fwiw I find gpt Sol on medium or low effort to also be faster than opus to the point of helping me not check out, but is quite smarter than dsv4 flash", "context": "", "use_case": "coding", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802093696, "item_id": 50130856, "url": "https://news.ycombinator.com/item?id=50130856", "by": "cjbgkagh", "date": "2026-10-17", "stance": "recommend", "quote": "I use Sol on medium as the work I'm doing now needs more memory than it needs intelligence and it's working great.", "excerpt": "The work I do cannot fit in 256k so it was just relearning the same things over and over again while making zero progress. AFAIK the 1M burns tokens at twice the rate but I use Sol on medium as the work I'm doing now needs more memory than it needs intelligence and it's working great. I use Opus 5.5 for the stuff that needs stronger intelligence.", "context": "", "use_case": "coding with large context", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802082050, "item_id": 50130128, "url": "https://news.ycombinator.com/item?id=50130128", "by": "clx75", "date": "2026-10-17", "stance": "neutral", "quote": "it should elevate to a higher level model (Sol), and ask its review", "excerpt": "Probably a week ago someone posted a thread with the title \"Why isn't the industry freaking out about DeepSeek 4.1 Flash?\" I have been intrigued about Chinese models for a long time, but never tried anything except an early Qwen. My daily driver was Luna, with occasional elevation to Terra or Sol - all on 5.6 as I found the output of 6 jarringly incomprehensible. I still had 95 USD on my OpenRouter account, so I thou", "context": "", "use_case": "plan review", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802077777, "item_id": 50129861, "url": "https://news.ycombinator.com/item?id=50129861", "by": "gregwebs", "date": "2026-10-17", "stance": "neutral", "quote": "I was trying out using Sol and liking the result", "excerpt": "The company's own news page: https://api-docs.deepseek.com/news/news261017 They are showing this new version scoring significantly higher across multiple coding benchmarks, beating Sol 5.6 (6.1 is the latest) across 10/11 benchmarks. I am going back to DeepSeek for writing code. I was trying out using Sol and liking the result, but this should close a lot of that gap. I use Opus/Sol for an implementation plan and cod", "context": "", "use_case": "planning and code review", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802068867, "item_id": 50129304, "url": "https://news.ycombinator.com/item?id=50129304", "by": "egeozcan", "date": "2026-10-17", "stance": "recommend", "quote": "Using sol 6.1 for review loops at the end.", "excerpt": "I use Opus for planning, through API on medium thinking, it doesn't cost anything significant and the rest can be happily done with deepseek or glm flash. Using sol 6.1 for review loops at the end. All running on pi. A very complex task tends to cost like 10$ max. It's comparable to a subscription with enormous flexibility. It's a steal.", "context": "", "use_case": "code review", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 800403120, "item_id": 50025195, "url": "https://news.ycombinator.com/item?id=50025195", "by": "aftbit", "date": "2026-10-09", "stance": "recommend", "quote": "Most of my work is with GPT-5.6-Sol.", "excerpt": "I have had good luck rotating between the three tiers of GPT 5.6 with occasional jumps up to Astra or Fable. Most of my work is with GPT-5.6-Sol. Simple tasks like data extraction or trivial refactors (rename this variable etc) I often push down to Terra or Luna, or even to self-hosted Gemma4:31B. Very tricky stuff, like planning a new feature, design review, or code review of a complex change across multiple reposit", "context": "", "use_case": "daily coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800204066, "item_id": 50012754, "url": "https://news.ycombinator.com/item?id=50012754", "by": "gregwebs", "date": "2026-10-08", "stance": "recommend", "quote": "It also doesn't come up with as good solutions as Opus/Sol.", "excerpt": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast. Previously I was always quickly running up to my Claude/Codex 5 hour window (on the $20/month plan). The cost savings of DeepSeek is real as shown in this article and I am using subsidized plans. DeepSeek is horrible at grilling sessions (the /grill* skills to make technical decisions). It doesn't know", "context": "", "use_case": "planning, review, coding", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800197698, "item_id": 50012356, "url": "https://news.ycombinator.com/item?id=50012356", "by": "apitman", "date": "2026-10-08", "stance": "recommend", "quote": "I use gpt-6.1-sol high for planning/orchestration", "excerpt": "> With my OpenCode Go sub of $10/month, DeepSeek is basically unlimited My OpenCode Go monthly window was scheduled to reset this morning. It was sitting at 22% used despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks (I use gpt-6.1-sol high for planning/orchestration). I had 1.5 hours left so I fired up first 10, then 20, and finally 50 concurrent subagents all working on re", "context": "", "use_case": "planning and orchestration", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800071281, "item_id": 50004455, "url": "https://news.ycombinator.com/item?id=50004455", "by": "zapnuk", "date": "2026-10-08", "stance": "neutral", "quote": "its about 10\u20ac for sol to plan or debug and 5\u20ac for luna to implement.", "excerpt": "At scale this make a difference. For me/us not that much. At my work we currently don't have a subscription and use API based billing. Some time ago the price for my usage was about 30-50\u20ac per day when we used opus for about everything. With luna+sol its about 10\u20ac for sol to plan or debug and 5\u20ac for luna to implement. At that point it doesn't quite matter to me if the cheap anthropic model is 20-50% more expensive co", "context": "uses API billing at work", "use_case": "planning and debugging", "switched_from": "", "story": "Claude Haiku 5.5"}, {"id": 800071154, "item_id": 50004447, "url": "https://news.ycombinator.com/item?id=50004447", "by": "jeremyjh", "date": "2026-10-08", "stance": "neutral", "quote": "Presently I'm using sol-high for the default agent which does orchestration", "excerpt": "I use oh-my-pi (omp.sh) - mostly with stock settings and skills but its a \"fully loaded\" harness that you don't really have to add anything to. I do change a couple of things: I set it to prefer subagents, and I enable rewind. It is crazy that rewind is not a default, it saves a TON of context - if the agent goes down a crazy path that burns a lot of tokens it can rewind to an earlier checkpoint with the exploration ", "context": "", "use_case": "orchestration, planning, review", "switched_from": "", "story": "Claude Haiku 5.5"}, {"id": 799988081, "item_id": 49999255, "url": "https://news.ycombinator.com/item?id=49999255", "by": "hbn", "date": "2026-10-07", "stance": "recommend", "quote": "now I'm on GPT-6 Sol. It's been a very good experience.", "excerpt": "What does it even mean to use \"Copilot\" these days? I use the Github Copilot app at work but it's really just a tool that gives models access to our Github repos and my local clones of them, and it's been super helpful to just tell it to reference other repos for patterns or integrations or whatever, get agents and an orchestrator to look at multiple repos to make a change across several projects and then implement t", "context": "", "use_case": "coding via Copilot", "switched_from": "Opus 4.6", "story": "Meta and Microsoft take steps to reduce employee usage of Claude AI"}, {"id": 799625137, "item_id": 49976571, "url": "https://news.ycombinator.com/item?id=49976571", "by": "chilmers", "date": "2026-10-06", "stance": "recommend", "quote": "I'm no longer hitting any of the capability limits I used to experience", "excerpt": "I have subscriptions to both, one for personal and one for work use, and I haven't really come close to hitting the limits since the release of Opus 5.5 and GPT-Sol 6.1, and I'm no longer hitting any of the capability limits I used to experience when using the \"mid\" tier models, so I don't miss using Astra. It feels like limits at pretty much a solved problem right now, for my level of use.", "context": "has subscriptions to both Anthropic and OpenAI", "use_case": "work and personal use", "switched_from": "", "story": "Anthropic Subscriptions Offer 5x+ More Value Than OpenAI"}, {"id": 799092848, "item_id": 49943303, "url": "https://news.ycombinator.com/item?id=49943303", "by": "Gareth321", "date": "2026-10-03", "stance": "recommend", "quote": "I used Sol 5.6 Extra High", "excerpt": "Elvis wasn't the one calling for censorship. It was religious viewers - the audience. I'm not denying that there were examples of creatives - like punks - who called for censorship. I'm describing a massive shift in scale and scope. I did as you suggested because it's important to have a broad and balanced view of the issues. I used the query \"Do you believe the scale and scope of creatives calling for censorship is ", "context": "", "use_case": "research queries", "switched_from": "", "story": "Newgrounds.com \u2013 A community of games, music, and art"}, {"id": 799092849, "item_id": 49943303, "url": "https://news.ycombinator.com/item?id=49943303", "by": "Gareth321", "date": "2026-10-03", "stance": "warn", "quote": "I'm deliberately avoiding Sol 6 right now due to some issues with hallucinations", "excerpt": "Elvis wasn't the one calling for censorship. It was religious viewers - the audience. I'm not denying that there were examples of creatives - like punks - who called for censorship. I'm describing a massive shift in scale and scope. I did as you suggested because it's important to have a broad and balanced view of the issues. I used the query \"Do you believe the scale and scope of creatives calling for censorship is ", "context": "", "use_case": "", "switched_from": "", "story": "Newgrounds.com \u2013 A community of games, music, and art"}, {"id": 798868403, "item_id": 49929275, "url": "https://news.ycombinator.com/item?id=49929275", "by": "pettijohn", "date": "2026-10-02", "stance": "neutral", "quote": "I typically ask GPT 6.1 Sol to review requirements", "excerpt": "I use PI WEB (https://pi-web.dev, there's more than one with that name) to orchestrate remotely. Pi lets me run local models (currently Qwen 3.8 Flash Next on Strix Halo 128GB) and flagship models (with subscription auth, not API pricing) side by side. I typically ask GPT 6.1 Sol to review requirements and then spawn a subsession with Qwen, review Qwen's work, ask Qwen to fix. About 95% success with a single pass lik", "context": "runs local models on Strix Halo 128GB", "use_case": "reviewing requirements and Qwen's work", "switched_from": "", "story": "Pi 1.0"}, {"id": 798569552, "item_id": 49910597, "url": "https://news.ycombinator.com/item?id=49910597", "by": "gertlabs", "date": "2026-09-30", "stance": "recommend", "quote": "6.1-Sol is also cheaper than Sonnet 5.5 and much smarter.", "excerpt": "We tested it across 100 unsaturated coding and engineering environments. Both Astra and 6.1-Sol are pretty comfortably ahead of Opus 5.5 in these types of evaluations, and both end up being cheaper than Opus via API usage. 6.1-Sol is also cheaper than Sonnet 5.5 and much smarter. The only verifiable domain Anthropic seems to be clearly ahead is chemistry (and perhaps also some unverifiable domains like being pleasant", "context": "", "use_case": "coding and engineering tasks", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798524307, "item_id": 49907769, "url": "https://news.ycombinator.com/item?id=49907769", "by": "yurishimo", "date": "2026-09-30", "stance": "recommend", "quote": "Sol for real problems", "excerpt": "Glad I'm not the only want to find it a bit janky/broken at times. They seem to constantly be pushing updates which is nice, but I treat it mostly as a black box. Anecdotally, I find the auto compaction (or what I assume is happening when the context magically drops) to be hit or miss. I do like how easy it is to use my work cursor sub and business chat gpt at the same time. Then I use nearly free cursor models for d", "context": "", "use_case": "harder coding problems", "switched_from": "", "story": "You said no MCP"}, {"id": 798483346, "item_id": 49905209, "url": "https://news.ycombinator.com/item?id=49905209", "by": "laurels-marts", "date": "2026-09-30", "stance": "warn", "quote": "OAI models simplify have no taste.", "excerpt": "100% in agreement. I pay for OAI sub and also use Codex exclusively at work for the past 8 months. I tried out fable 5.1 the day it was released and coming from gpt-5.6-sol I was truly mind blown (both in terms of code and prose it was generating - outputs I could finally enjoy reading and looking at). Then when opus 5.5 came out, again same thing + far cheaper and faster. I went from using OAI exclusively the entire", "context": "", "use_case": "coding", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798446625, "item_id": 49902914, "url": "https://news.ycombinator.com/item?id=49902914", "by": "Benjamin_Dobell", "date": "2026-09-30", "stance": "neutral", "quote": "model = \"gpt-6.1-sol\"", "excerpt": "The context window is configurable. I've been using ~600k for months. No, not API pricing, on a Codex sub. ~/.codex/config.toml model = \"gpt-6.1-sol\" model_context_window = 700000 model_auto_compact_token_limit = 630000", "context": "", "use_case": "coding via Codex", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798439075, "item_id": 49902442, "url": "https://news.ycombinator.com/item?id=49902442", "by": "srmckee", "date": "2026-09-29", "stance": "warn", "quote": "I was using Sol originally which was ok but also irritating because it would mess up and make a buggy ui.", "excerpt": "Hey HN! Spencer here - wanted to share a token compression tool that I\u2019ve built for myself to save 30% costs on codex! I've been building a ton of web apps to replace saas i'm too cheap to keep paying for like chatgpt, notion, myfitnesspal, etc. and so i can get the exact features i want (+ no ads!). I was using Sol originally which was ok but also irritating because it would mess up and make a buggy ui. so I switche", "context": "", "use_case": "coding model", "switched_from": "", "story": "Show HN: Free token compression CLI, saves codex bills by 30%"}, {"id": 798397648, "item_id": 49899853, "url": "https://news.ycombinator.com/item?id=49899853", "by": "jorblumesea", "date": "2026-09-29", "stance": "recommend", "quote": "it's solid and cheaper", "excerpt": "yeah I use sol constantly and have done maybe $15 of spend in the past week. it's solid and cheaper. this is at least 4-5 investigations, prs, whatever per day.", "context": "", "use_case": "investigations and PRs", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798364464, "item_id": 49897779, "url": "https://news.ycombinator.com/item?id=49897779", "by": "ChaseRensberger", "date": "2026-09-29", "stance": "recommend", "quote": "I expect this will be my new daily driver.", "excerpt": "very surprised by the sentiment against GPT 6.0 Sol, I've been using it exclusively since release and it feels like a cheaper astra to me. admittedly I haven't tried any anthropic models in a while other than small tests since i can't use my anthropic subscription in other harnesses (like OpenAI has supported natively for a long time). If OpenAI cuts alternative harness support it will be a weird day trying to figure", "context": "", "use_case": "daily driver model", "switched_from": "GPT 6 Sol", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798364465, "item_id": 49897779, "url": "https://news.ycombinator.com/item?id=49897779", "by": "ChaseRensberger", "date": "2026-09-29", "stance": "recommend", "quote": "I've been using it exclusively since release and it feels like a cheaper astra to me.", "excerpt": "very surprised by the sentiment against GPT 6.0 Sol, I've been using it exclusively since release and it feels like a cheaper astra to me. admittedly I haven't tried any anthropic models in a while other than small tests since i can't use my anthropic subscription in other harnesses (like OpenAI has supported natively for a long time). If OpenAI cuts alternative harness support it will be a weird day trying to figure", "context": "", "use_case": "daily model", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798302642, "item_id": 49893915, "url": "https://news.ycombinator.com/item?id=49893915", "by": "dools", "date": "2026-09-29", "stance": "neutral", "quote": "Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt", "excerpt": "I go for weeks without topping up my Deepseek balance, and have to top up my OpenAI balance at least every week. GPT 6 Luna closed the gap significantly for sure (it seems to be about twice as expensive as DS v4.1 Flash), but Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do. Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt, ", "context": "", "use_case": "hard problems fallback", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 798274050, "item_id": 49892128, "url": "https://news.ycombinator.com/item?id=49892128", "by": "joshstrange", "date": "2026-09-29", "stance": "warn", "quote": "I burned through my whole week in 1 day on Sol.", "excerpt": "I can believe that (20x is not 4x of the 5x plan) but I\u2019ve been on the $100/mo Codex and the $200/mo Claude plans and I can run the whole week with multiple agents on Claude, mean while I burned through my whole week in 1 day on Sol. Yes, it\u2019s half the price of my Claude plan but even if OpenAi actually gave 4x usage for the next plan up (which they just halved), it would still be less than the usage I get from Claud", "context": "", "use_case": "coding", "switched_from": "", "story": "OpenAI: Tomorrow we are re-opening the Pro $200 subscription"}, {"id": 798177712, "item_id": 49886107, "url": "https://news.ycombinator.com/item?id=49886107", "by": "cromka", "date": "2026-09-28", "stance": "warn", "quote": "This is astonishingly bad and it is nowhere close to what Sol 5.6 was a month ago.", "excerpt": "Not even GPT6 Sol cannot match DeepSeek 4.1 in my work, with outrageous bugs. Don't get me started on Luna. From my today's session with Sol: - it actually failed to correctly understand a simple English grammar and logical implication of it, then when challenged it admitted its mistake but couldn't explain why it made it. - for the code I am working on, I asked to create two PRs for the two small features (couple li", "context": "", "use_case": "coding", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 796936434, "item_id": 49808527, "url": "https://news.ycombinator.com/item?id=49808527", "by": "jeremyjh", "date": "2026-09-22", "stance": "recommend", "quote": "Sol high as the orchestrator, Sol max as Planner & Reviewer", "excerpt": "You can get a lot of Codex usage out of that same sub on top of ChatGPT usage. Its a really good value and you can use that sub in any harness. In OMP I have Sol high as the orchestrator, Sol max as Planner & Reviewer, Luna max as task/coder. Very good setup. I'm on pro now and there are weekends when I use half a week's usage but I'll have 5 or 6 sessions going at once for many hours each day.", "context": "", "use_case": "orchestration, planning, review", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 796902128, "item_id": 49806383, "url": "https://news.ycombinator.com/item?id=49806383", "by": "apitman", "date": "2026-09-22", "stance": "recommend", "quote": "gpt-5.6-sol high has been my go-to for months.", "excerpt": "Similar for me. gpt-5.6-sol high has been my go-to for months. One of the reasons I'm pushing myself to try open models more is because it lends some level of guarantee I can continue to use the same tool as long as I want to. And I think we may just be getting to the point the open models are >= 5.6 Sol for coding.", "context": "", "use_case": "coding", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 795344272, "item_id": 49709017, "url": "https://news.ycombinator.com/item?id=49709017", "by": "trencedamp", "date": "2026-09-15", "stance": "warn", "quote": "it consistently, confidently failed", "excerpt": "It's not the word I'm taking issue with, it's the concept. From your post you are surprised that the model makes mistakes and does not recognize things, and these are only surprising if you imagine the models to be thinking about things. We can argue about words like thinking and intelligence all day long, and so can an LLM. But at the end of the day the LLM is only mimicking the processes you or I use. Last week I t", "context": "", "use_case": "counting letters in a long string", "switched_from": "", "story": "Pion, an agent designed to run any company autonomously"}, {"id": 795268496, "item_id": 49704281, "url": "https://news.ycombinator.com/item?id=49704281", "by": "glimshe", "date": "2026-09-14", "stance": "recommend", "quote": "Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.", "excerpt": "I select my review model based on change complexity. If I'm confident that the change is localized (given that I always maintain an up to date and complete mental model of the application), I'll use Luna/Terra. For deeper changes, I use Sol/Astra. Sol is so good for code reviews that I'll only engage Astra in the very riskiest changes.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}]}