{"slug": "openai-api", "name": "OpenAI API", "category": "ai", "category_name": "AI tools and apps", "description": "", "recommenders": 9, "commenters": 57, "warners": 13, "score": 5.834, "rising": 1.04, "first": "2025-11-06", "latest": "2026-10-14", "why": "This was so bad that I switched to OpenAI API instead.", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 9, "neutral": 34, "compare": 2, "warn": 13}, "spark": [0, 0, 1, 3, 0, 0, 1, 0, 1, 0, 1, 2], "listed": true, "loved": [], "caveat": "", "short_quote": "This was so bad that I switched to OpenAI API instead.", "kind": "", "icon": true, "domain": "openai.com", "github_repo": null, "official_url": "https://openai.com/api/", "pricing": "", "aliases": ["OpenAI API", "OpenAI", "openai"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 801357265, "item_id": 50084829, "url": "https://news.ycombinator.com/item?id=50084829", "by": "sunaookami", "date": "2026-10-14", "version": null, "stance": "recommend", "quote": "This was so bad that I switched to OpenAI API instead.", "excerpt": "Huh? The constant failures in the Gemini API and overnight changes are well documented, there was a time from January to roughly July (maybe it still persists) where the API constantly spew 503 errors. This was so bad that I switched to OpenAI API instead. See e.g. https://discuss.ai.google.dev/t/frequent-503-errors-service-unavailable-across-all-models/116450", "context": "", "use_case": "LLM API", "switched_from": "Gemini API", "story": "Gemini 4 Argon"}, {"id": 800733937, "item_id": 50045871, "url": "https://news.ycombinator.com/item?id=50045871", "by": "Topfi", "date": "2026-10-11", "version": null, "stance": "recommend", "quote": "Did a rerun with the official OpenAI API directly and was able to reproduce the issue", "excerpt": "Thanks for the warning. Did a rerun with the official OpenAI API directly and was able to reproduce the issue on those queries, so at least in this case, not OpenRouters fault. Also, shared them with Brian via Mail [0], hopefully useful to them. Do occasionally cross-compare official with OpenRouter and have yet to encounter anything that wouldn't be replaced by run-to-run variance, but will keep an eye out for it, t", "context": "", "use_case": "Model API", "switched_from": "", "story": "Decisions API is in public beta"}, {"id": 797686528, "item_id": 49855408, "url": "https://news.ycombinator.com/item?id=49855408", "by": "jychang", "date": "2026-09-26", "version": null, "stance": "recommend", "quote": "OpenAI\u2019s API format has become an industry standard.", "excerpt": "> The Muse public APIs seem to be heavily inspired by OpenAI's, they even support the richer \"responses\" API. ??? Every single provider basically copy openai's api endpoint https://api.openai.com/v1/chat/completions https://router.huggingface.co/v1/chat/completions https://integrate.api.nvidia.com/v1/chat/completions https://generativelanguage.googleapis.com/v1beta/openai/chat/completions https://api.x.ai/v1/chat/com", "context": "", "use_case": "", "switched_from": "", "story": "Meta's Muse appears to use an OpenAI model labeled muse-special"}, {"id": 781254897, "item_id": 48828431, "url": "https://news.ycombinator.com/item?id=48828431", "by": "deadlast2", "date": "2026-07-07", "version": null, "stance": "recommend", "quote": "Open AI gives much more value for money than Claude", "excerpt": "I use Open AI and Claude a lot right like a lot everyday for hours multiple hours. Open AI gives much more value for money than Claude much more I'd say x 10. Mainly I use it for writing fiction books and literally Claude is locked 90% of everyday trying to jip me for tokens. It's not as good at coding for what I do which is a very complicated application. However it is very good at writing it's really good which is ", "context": "writes fiction books with AI", "use_case": "writing fiction books", "switched_from": "", "story": "GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday"}, {"id": 769414033, "item_id": 48088377, "url": "https://news.ycombinator.com/item?id=48088377", "by": "franze", "date": "2026-05-10", "version": null, "stance": "recommend", "quote": "standardized API is openai api so thats what i went with", "excerpt": "i researched that question for apfel https://github.com/Arthur-Ficial/apfel and standardized API is openai api so thats what i went with", "context": "", "use_case": "", "switched_from": "", "story": "Local AI needs to be the norm"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 794067696, "item_id": 49629231, "url": "https://news.ycombinator.com/item?id=49629231", "by": "burkaman", "date": "2026-09-09", "version": null, "stance": "warn", "quote": "it seems more likely than not that at least some amount of his usage became training data", "excerpt": "I don't understand how they can say it's unlikely. It's objectively true that they train on de-identified user data (https://openai.com/policies/how-your-data-is-used-to-improve-model-performance/), and objectively true that they encourage users to submit such data (the setting is on by default). Since it's de-identified I can believe that they can't give a straight yes or no answer here, but it seems more likely tha", "context": "", "use_case": "", "switched_from": "", "story": "Navier-Stokes \u2013 Tristan Buckmaster [pdf]"}, {"id": 777897696, "item_id": 48618606, "url": "https://news.ycombinator.com/item?id=48618606", "by": "JimDabell", "date": "2026-06-21", "version": null, "stance": "warn", "quote": "What is especially bad is that if you fail the verification process, they won\u2019t let you retry", "excerpt": "OpenAI also has this kind of check. What is especially bad is that if you fail the verification process, they won\u2019t let you retry \u2013 you are permanently locked out from the top models. They aren\u2019t clear about this upfront during the process, so make sure the lighting is good when you scan your ID! https://help.openai.com/en/articles/10910291-api-organization-verification", "context": "", "use_case": "Model access via API", "switched_from": "", "story": "Anthropic to Require ID Verification for Certain Capabilities Starting July 8"}, {"id": 772694976, "item_id": 48293436, "url": "https://news.ycombinator.com/item?id=48293436", "by": "weird-eye-issue", "date": "2026-05-27", "version": null, "stance": "warn", "quote": "China and Hong Kong are conveniently missing", "excerpt": "https://developers.openai.com/api/docs/supported-countries You can check for yourself here to see that China and Hong Kong are conveniently missing. We do see blocking from Anthropic and Gemini as well in some regions Also even though Vietnam and the Philippines are technically supported we do see blocking from some IP addresses in those regions too", "context": "", "use_case": "", "switched_from": "", "story": "Xiaomi MiMo-v2.5 Series API Permanent Price Reduction Up to 99%"}, {"id": 760390496, "item_id": 47524406, "url": "https://news.ycombinator.com/item?id=47524406", "by": "Ucalegon", "date": "2026-03-25", "version": null, "stance": "warn", "quote": "OAI does not have regional processing for AU", "excerpt": "The devil is in the details. For example, OAI does not have regional processing for AU [0] and their ZDR does not cover files[1]. Anthropic's ZDR [2] also does not cover files, so you really need to be careful, as a patient/consumer, to ensure that your health, or other sensitive data, that is being processed by SaaS frontier models is not contained in files. Which is asking a a lot of the medical provider to know ho", "context": "", "use_case": "", "switched_from": "", "story": "Health NZ staff told to stop using ChatGPT to write clinical notes"}]}], "who": [{"by": "deadlast2", "context": "writes fiction books with AI", "url": "https://news.ycombinator.com/item?id=48828431", "date": "2026-07-07"}, {"by": "dmc258", "context": "biomedical researcher", "url": "https://news.ycombinator.com/item?id=47381219", "date": "2026-03-14"}, {"by": "vincent_s", "context": "full stack developer and AI engineer, 15+ YoE, seeking work", "url": "https://news.ycombinator.com/item?id=46476500", "date": "2026-01-03"}, {"by": "gremlinsinc", "context": "Full-stack developer, 12+ years, seeking contract work", "url": "https://news.ycombinator.com/item?id=46114447", "date": "2025-12-01"}, {"by": "markvitals", "context": "built the project", "url": "https://news.ycombinator.com/item?id=46053309", "date": "2025-11-25"}], "tags": [], "use_cases": ["production llm apis", "llm api access", "llm api", "model api", "ai steps for email grouping", "agent/model access"], "compared": [{"slug": "claude", "name": "Claude", "co": 25, "switched_from": 0, "switched_to": 2}, {"slug": "gemini", "name": "Gemini", "co": 13, "switched_from": 0, "switched_to": 1}, {"slug": "next-js", "name": "Next.js", "co": 9, "switched_from": 0, "switched_to": 0}, {"slug": "claude-code", "name": "Claude Code", "co": 8, "switched_from": 0, "switched_to": 0}, {"slug": "postgresql", "name": "PostgreSQL", "co": 7, "switched_from": 0, "switched_to": 0}, {"slug": "gemini-api", "name": "Gemini API", "co": 3, "switched_from": 1, "switched_to": 1}, {"slug": "react", "name": "React", "co": 6, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49187575", "title": "Muse Code and Muse Spark 1.2", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=48357725", "title": "Ask HN: Who is hiring? (June 2026)", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=46223311", "title": "Getting a Gemini API key is an exercise in frustration", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=46108940", "title": "Ask HN: Who wants to be hired? (December 2025)", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=45840088", "title": "You should write an agent", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=50073879", "title": "Gemini 4 Argon", "mentions": 1}], "sentiment_by_month": {"2025-11": {"neutral": 4}, "2025-12": {"neutral": 6, "warn": 3}, "2026-01": {"warn": 3, "neutral": 3, "recommend": 1}, "2026-02": {"neutral": 3, "recommend": 3, "warn": 2}, "2026-03": {"neutral": 3, "warn": 1}, "2026-04": {"neutral": 4, "compare": 1}, "2026-05": {"neutral": 3, "recommend": 1, "warn": 1}, "2026-06": {"neutral": 2, "warn": 1}, "2026-07": {"neutral": 2, "recommend": 1}, "2026-08": {"compare": 1, "neutral": 1, "warn": 1}, "2026-09": {"warn": 1, "neutral": 1, "recommend": 1}, "2026-10": {"neutral": 2, "recommend": 2}}, "versions": [], "version_unnamed": 0, "family_ref": null, "runs_on": null, "lines": [], "quotes": [{"id": 801357265, "item_id": 50084829, "url": "https://news.ycombinator.com/item?id=50084829", "by": "sunaookami", "date": "2026-10-14", "version": null, "stance": "recommend", "quote": "This was so bad that I switched to OpenAI API instead.", "excerpt": "Huh? The constant failures in the Gemini API and overnight changes are well documented, there was a time from January to roughly July (maybe it still persists) where the API constantly spew 503 errors. This was so bad that I switched to OpenAI API instead. See e.g. https://discuss.ai.google.dev/t/frequent-503-errors-service-unavailable-across-all-models/116450", "context": "", "use_case": "LLM API", "switched_from": "Gemini API", "story": "Gemini 4 Argon"}, {"id": 800733937, "item_id": 50045871, "url": "https://news.ycombinator.com/item?id=50045871", "by": "Topfi", "date": "2026-10-11", "version": null, "stance": "recommend", "quote": "Did a rerun with the official OpenAI API directly and was able to reproduce the issue", "excerpt": "Thanks for the warning. Did a rerun with the official OpenAI API directly and was able to reproduce the issue on those queries, so at least in this case, not OpenRouters fault. Also, shared them with Brian via Mail [0], hopefully useful to them. Do occasionally cross-compare official with OpenRouter and have yet to encounter anything that wouldn't be replaced by run-to-run variance, but will keep an eye out for it, t", "context": "", "use_case": "Model API", "switched_from": "", "story": "Decisions API is in public beta"}, {"id": 800628880, "item_id": 50039305, "url": "https://news.ycombinator.com/item?id=50039305", "by": "TomatoCo", "date": "2026-10-11", "version": null, "stance": "neutral", "quote": "OpenAI and Anthropic are both 10/mil in.", "excerpt": "Where are you getting 150-400 per million input? https://developers.openai.com/api/docs/pricing https://platform.claude.com/docs/en/about-claude/pricing OpenAI and Anthropic are both 10/mil in. https://openrouter.ai/z-ai/glm-5.3#providers https://openrouter.ai/moonshotai/kimi-k3#providers Other frontier models are like 1-3/mil in. Also, 500 billion is 500,000 millions. At the lower end of your 140/mil estimate that's", "context": "", "use_case": "", "switched_from": "", "story": "500B tokens later: Letting AI agents decompile a first-person shooter"}, {"id": 798726129, "item_id": 49920383, "url": "https://news.ycombinator.com/item?id=49920383", "by": "jtiret", "date": "2026-10-01", "version": null, "stance": "neutral", "quote": "We're using the standard OpenAI API setup", "excerpt": "Good questions! For some AI steps, we send sender/recipient information, subjects and truncated message bodies from a sample of messages (typically between 100 and 300 depending on the step). For instance we use that to find additional groups of emails to archive, suggest filters to skip the inbox and move messages under a label or identify newsletters. A local tool/BYOK version could definitely make sense. We didn't", "context": "", "use_case": "AI steps for email grouping", "switched_from": "", "story": "Show HN: Inbox Cleanup \u2013 get to inbox zero in Gmail"}, {"id": 797686528, "item_id": 49855408, "url": "https://news.ycombinator.com/item?id=49855408", "by": "jychang", "date": "2026-09-26", "version": null, "stance": "recommend", "quote": "OpenAI\u2019s API format has become an industry standard.", "excerpt": "> The Muse public APIs seem to be heavily inspired by OpenAI's, they even support the richer \"responses\" API. ??? Every single provider basically copy openai's api endpoint https://api.openai.com/v1/chat/completions https://router.huggingface.co/v1/chat/completions https://integrate.api.nvidia.com/v1/chat/completions https://generativelanguage.googleapis.com/v1beta/openai/chat/completions https://api.x.ai/v1/chat/com", "context": "", "use_case": "", "switched_from": "", "story": "Meta's Muse appears to use an OpenAI model labeled muse-special"}, {"id": 794415280, "item_id": 49650955, "url": "https://news.ycombinator.com/item?id=49650955", "by": "TZubiri", "date": "2026-09-10", "version": null, "stance": "neutral", "quote": "We use the OAI API because there is no local equivalent", "excerpt": "Why would you prefer to use the API if you can have something running locally? We use the OAI API because there is no local equivalent, I'm assuming this is just the codex client running on the cloud?", "context": "", "use_case": "agent/model access", "switched_from": "", "story": "OpenAI Agents API"}, {"id": 794067696, "item_id": 49629231, "url": "https://news.ycombinator.com/item?id=49629231", "by": "burkaman", "date": "2026-09-09", "version": null, "stance": "warn", "quote": "it seems more likely than not that at least some amount of his usage became training data", "excerpt": "I don't understand how they can say it's unlikely. It's objectively true that they train on de-identified user data (https://openai.com/policies/how-your-data-is-used-to-improve-model-performance/), and objectively true that they encourage users to submit such data (the setting is on by default). Since it's de-identified I can believe that they can't give a straight yes or no answer here, but it seems more likely tha", "context": "", "use_case": "", "switched_from": "", "story": "Navier-Stokes \u2013 Tristan Buckmaster [pdf]"}, {"id": 789543616, "item_id": 49346476, "url": "https://news.ycombinator.com/item?id=49346476", "by": "Telaneo", "date": "2026-08-18", "version": null, "stance": "warn", "quote": "", "excerpt": "https://community.e.foundation/t/voice-to-text-feature-using-open-ai/70509", "context": "", "use_case": "", "switched_from": "", "story": "Fairphone is now officially available in the United States"}, {"id": 787066608, "item_id": 49191663, "url": "https://news.ycombinator.com/item?id=49191663", "by": "no-name-here", "date": "2026-08-05", "version": null, "stance": "neutral", "quote": "I enabled all data sharing settings", "excerpt": "I enabled all data sharing settings but still don\u2019t have a message about free use on that screen - the help page says free tokens are available to \u201csome\u201d users - is that 1% of users, 40% of users, etc? Does your screen have the message that you are receiving free tokens? https://platform.openai.com/settings/organization/data-controls/sharing", "context": "", "use_case": "", "switched_from": "", "story": "Muse Code and Muse Spark 1.2"}, {"id": 787054912, "item_id": 49190932, "url": "https://news.ycombinator.com/item?id=49190932", "by": "moonu", "date": "2026-08-05", "version": null, "stance": "compare", "quote": "OpenAI, for a while now, has had free API usage when data sharing is enabled", "excerpt": "I've been surprised by the reception to this, as OpenAI, for a while now, has had free API usage when data sharing is enabled (https://help.openai.com/en/articles/10306912-sharing-feedback-evaluation-and-fine-tuning-data-and-api-inputs-and-outputs-with-openai)", "context": "", "use_case": "", "switched_from": "", "story": "Muse Code and Muse Spark 1.2"}, {"id": 781343089, "item_id": 48833943, "url": "https://news.ycombinator.com/item?id=48833943", "by": "catalystwayfare", "date": "2026-07-08", "version": null, "stance": "neutral", "quote": "Stack: Claude, OpenAI, Gemini, and Perplexity APIs in production", "excerpt": "Catalyst\u00b7Wayfare AI | Lead AI Engineer + Agent Builders | Remote (US) | Full-time + Contract | https://www.catalystwayfare.ai/careers We build and ship production AI systems for mid-market enterprises in regulated and technical domains \u2014 finance, energy, industrials, healthcare. We do not stop at PowerPoint AI strategy. We ship working systems alongside our clients' engineering teams. Small teams, deep expertise, rea", "context": "", "use_case": "Production LLM APIs", "switched_from": "", "story": "Ask HN: Who is hiring? (July 2026)"}, {"id": 781254897, "item_id": 48828431, "url": "https://news.ycombinator.com/item?id=48828431", "by": "deadlast2", "date": "2026-07-07", "version": null, "stance": "recommend", "quote": "Open AI gives much more value for money than Claude", "excerpt": "I use Open AI and Claude a lot right like a lot everyday for hours multiple hours. Open AI gives much more value for money than Claude much more I'd say x 10. Mainly I use it for writing fiction books and literally Claude is locked 90% of everyday trying to jip me for tokens. It's not as good at coding for what I do which is a very complicated application. However it is very good at writing it's really good which is ", "context": "writes fiction books with AI", "use_case": "writing fiction books", "switched_from": "", "story": "GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday"}, {"id": 780259062, "item_id": 48766191, "url": "https://news.ycombinator.com/item?id=48766191", "by": "emanuelecz", "date": "2026-07-02", "version": null, "stance": "neutral", "quote": "Claude + OpenAI APIs", "excerpt": "Location: Basque Country, Spain (CET/CEST) Remote: Yes, remote only. EU contract, or US via EOR (Deel / Remote.com). Registered freelancer, can also invoice B2B. Willing to relocate: No Technologies: Python, LangGraph/LangChain, RAG (hybrid BM25 + dense retrieval, VoyageAI embeddings + reranking, ChromaDB, pgvector), LLM evals (golden sets, LLM-as-judge, judge meta-evaluation), Claude + OpenAI APIs, FastAPI, PostgreS", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: Who wants to be hired? (July 2026)"}, {"id": 777897696, "item_id": 48618606, "url": "https://news.ycombinator.com/item?id=48618606", "by": "JimDabell", "date": "2026-06-21", "version": null, "stance": "warn", "quote": "What is especially bad is that if you fail the verification process, they won\u2019t let you retry", "excerpt": "OpenAI also has this kind of check. What is especially bad is that if you fail the verification process, they won\u2019t let you retry \u2013 you are permanently locked out from the top models. They aren\u2019t clear about this upfront during the process, so make sure the lighting is good when you scan your ID! https://help.openai.com/en/articles/10910291-api-organization-verification", "context": "", "use_case": "Model access via API", "switched_from": "", "story": "Anthropic to Require ID Verification for Certain Capabilities Starting July 8"}, {"id": 774162337, "item_id": 48385146, "url": "https://news.ycombinator.com/item?id=48385146", "by": "catalystwayfare", "date": "2026-06-03", "version": null, "stance": "neutral", "quote": "Claude + OpenAI + Gemini + Perplexity APIs in production", "excerpt": "Catalyst\u00b7Wayfare AI | Lead AI Engineer + Agent Builder + Platform Engineer | Remote (US) | Full-time / Part-time / Contract | https://catalystwayfare.ai/careers AI transformation firm shipping production AI systems for mid-market enterprises in regulated and technical domains \u2014 finance, energy, industrials, healthcare. We do not stop at PowerPoint AI strategy. We ship working systems alongside our clients' engineerin", "context": "", "use_case": "Production LLM APIs", "switched_from": "", "story": "Ask HN: Who is hiring? (June 2026)"}, {"id": 773791953, "item_id": 48361997, "url": "https://news.ycombinator.com/item?id=48361997", "by": "kuznetsov_va", "date": "2026-06-01", "version": null, "stance": "neutral", "quote": "production calls OpenAI's API from Python", "excerpt": "Veedma | Founding Product Engineer + Founding Growth Engineer | Remote (US) | Full-time | $150-200k + 1.0-1.25% equity | https://app.dover.com/jobs/veedma Veedma is a telehealth practice for men's hormonal and sexual health. We treat low testosterone, ED, and vitamin D deficiency in men 18 and up, fully online. Cash-pay subscription covers consultations, lab panels twice yearly, prescribed medications, and unlimited ", "context": "", "use_case": "Production LLM calls", "switched_from": "", "story": "Ask HN: Who is hiring? (June 2026)"}, {"id": 773318577, "item_id": 48332411, "url": "https://news.ycombinator.com/item?id=48332411", "by": "red_hare", "date": "2026-05-29", "version": null, "stance": "neutral", "quote": "OpenAI API also supports defer_loading", "excerpt": "OpenAI API also supports defer_loading https://developers.openai.com/api/docs/guides/tools-tool-search And it's not actually necessary for it to exist at the API level. It's a pattern. Making it API-side is just an optimization. To do it client-side: 1. Define a single tool, tool_search 2. List the names of your deferred tools in context (or tool_search's description) 3. When tool_search is called, match the query ag", "context": "", "use_case": "", "switched_from": "", "story": "MCP Is Dead"}, {"id": 772694976, "item_id": 48293436, "url": "https://news.ycombinator.com/item?id=48293436", "by": "weird-eye-issue", "date": "2026-05-27", "version": null, "stance": "warn", "quote": "China and Hong Kong are conveniently missing", "excerpt": "https://developers.openai.com/api/docs/supported-countries You can check for yourself here to see that China and Hong Kong are conveniently missing. We do see blocking from Anthropic and Gemini as well in some regions Also even though Vietnam and the Philippines are technically supported we do see blocking from some IP addresses in those regions too", "context": "", "use_case": "", "switched_from": "", "story": "Xiaomi MiMo-v2.5 Series API Permanent Price Reduction Up to 99%"}, {"id": 769414033, "item_id": 48088377, "url": "https://news.ycombinator.com/item?id=48088377", "by": "franze", "date": "2026-05-10", "version": null, "stance": "recommend", "quote": "standardized API is openai api so thats what i went with", "excerpt": "i researched that question for apfel https://github.com/Arthur-Ficial/apfel and standardized API is openai api so thats what i went with", "context": "", "use_case": "", "switched_from": "", "story": "Local AI needs to be the norm"}, {"id": 768782833, "item_id": 48048927, "url": "https://news.ycombinator.com/item?id=48048927", "by": "theshrike79", "date": "2026-05-07", "version": null, "stance": "neutral", "quote": "uses a web page + phone camera + OpenAI API to identify physical DVDs", "excerpt": "> When you\u2019re bored, your mind goes to places it wouldn\u2019t otherwise go. Curiosity kicks in. Curiosity is a precursor to learning. Learning engages the brain and is fun. But it\u2019s not fun all the time, some of it is challenging and frustrating (which is good, that\u2019s the process that teaches you). And I love how I can go from a curious brainfart \"hmm, could I do a movie catalogue app that uses a web page + phone camera ", "context": "", "use_case": "identifying DVD covers", "switched_from": "", "story": "Vibe coding and agentic engineering are getting closer than I'd like"}, {"id": 767610169, "item_id": 47975635, "url": "https://news.ycombinator.com/item?id=47975635", "by": "knbrlo", "date": "2026-05-01", "version": null, "stance": "neutral", "quote": "Technologies: Go, TypeScript, AWS, Terraform, SpiceDB, Stripe Connect, PostgreSQL, React, React Native, Claude/OpenAI APIs, Eleven Labs", "excerpt": "Location: - Lehi, Utah, USA - Remote: Yes, remote-only - Willing to relocate: No - Technologies: Go, TypeScript, AWS, Terraform, SpiceDB, Stripe Connect, PostgreSQL, React, React Native, Claude/OpenAI APIs, Eleven Labs 20+ years shipping production SaaS. Currently CTO of Allset (AI-native ops for construction trades) \u2014 multi-tenant, Stripe Connect, mobile + web; just shipped to AWS this week. Before Allset: 10 years ", "context": "", "use_case": "AI APIs", "switched_from": "", "story": "Ask HN: Who wants to be hired? (May 2026)"}, {"id": 766152272, "item_id": 47884517, "url": "https://news.ycombinator.com/item?id=47884517", "by": "kiratp", "date": "2026-04-23", "version": null, "stance": "compare", "quote": "OpenAI does this for all API calls", "excerpt": "OpenAI does this for all API calls > Our systems will smartly ignore any reasoning items that aren\u2019t relevant to your functions, and only retain those in context that are relevant. You can pass reasoning items from previous responses either using the previous_response_id parameter, or by manually passing in all the output items from a past response into the input of a new one. https://developers.openai.com/api/docs/g", "context": "works on AI at Microsoft", "use_case": "", "switched_from": "", "story": "An update on recent Claude Code quality reports"}, {"id": 764681297, "item_id": 47792581, "url": "https://news.ycombinator.com/item?id=47792581", "by": "imafish", "date": "2026-04-16", "version": null, "stance": "neutral", "quote": "OpenAI also worked like this last time I used it", "excerpt": "OpenAI also worked like this last time I used it - not sure if that's changed.", "context": "", "use_case": "", "switched_from": "", "story": "\u20ac54k spike in 13h from unrestricted Firebase browser key accessing Gemini APIs"}, {"id": 764067584, "item_id": 47754224, "url": "https://news.ycombinator.com/item?id=47754224", "by": "jwr", "date": "2026-04-13", "version": null, "stance": "neutral", "quote": "I use the OpenAI API for everything", "excerpt": "I use the OpenAI API for everything. I think codex is more polished, but I don't really prefer anything: I haven't used them enough. I mostly use Claude Code.", "context": "", "use_case": "", "switched_from": "", "story": "I ran Gemma 4 as a local model in Codex CLI"}, {"id": 762333040, "item_id": 47645815, "url": "https://news.ycombinator.com/item?id=47645815", "by": "EnPissant", "date": "2026-04-04", "version": null, "stance": "neutral", "quote": "Not true.", "excerpt": "Not true. https://developers.openai.com/api/docs/guides/your-data", "context": "", "use_case": "", "switched_from": "", "story": "A case study in testing with 100 of Claude agents in parallel"}, {"id": 762095906, "item_id": 47630994, "url": "https://news.ycombinator.com/item?id=47630994", "by": "badmonday", "date": "2026-04-03", "version": null, "stance": "neutral", "quote": "Built with Next.js and the OpenAI API.", "excerpt": "XKCD #936 made the case for passphrases over random character soup back in 2011, and the argument is even stronger now. I wanted a generator that actually follows the logic of that comic \u2014 real words, memorable phrases, easy to type. The twist: it uses AI to generate nonsensical but grammatically plausible phrases (so you don't get \"horse battery staple correct\" clones), then gives you 6 variants of the same phrase \u2014", "context": "", "use_case": "", "switched_from": "", "story": "I built an Xkcd #936-style passphrase generator"}, {"id": 760390496, "item_id": 47524406, "url": "https://news.ycombinator.com/item?id=47524406", "by": "Ucalegon", "date": "2026-03-25", "version": null, "stance": "warn", "quote": "OAI does not have regional processing for AU", "excerpt": "The devil is in the details. For example, OAI does not have regional processing for AU [0] and their ZDR does not cover files[1]. Anthropic's ZDR [2] also does not cover files, so you really need to be careful, as a patient/consumer, to ensure that your health, or other sensitive data, that is being processed by SaaS frontier models is not contained in files. Which is asking a a lot of the medical provider to know ho", "context": "", "use_case": "", "switched_from": "", "story": "Health NZ staff told to stop using ChatGPT to write clinical notes"}, {"id": 758099506, "item_id": 47381219, "url": "https://news.ycombinator.com/item?id=47381219", "by": "dmc258", "date": "2026-03-14", "version": null, "stance": "neutral", "quote": "OpenAI API", "excerpt": "Hi HN, I built a small tool that helps prepare journal club presentations from research papers. In many labs, preparing a journal club involves: reading the paper summarizing key findings identifying strengths/limitations generating discussion questions building slides That process can take several hours. I wanted to see how much of that workflow could be automated while still producing something useful for scientifi", "context": "biomedical researcher", "use_case": "LLM analysis of paper chunks", "switched_from": "", "story": null}, {"id": 755719652, "item_id": 47232478, "url": "https://news.ycombinator.com/item?id=47232478", "by": "rgb1903", "date": "2026-03-03", "version": null, "stance": "neutral", "quote": "Built with Next.js, Firebase Admin SDK, and OpenAI (gpt-4o-mini", "excerpt": "Hi HN, I built PrecisionAudit because I got tired of the default \"turd polishing\" reflex of Large Language Models. Whenever you ask ChatGPT to improve landing page copy or a cold email, its default behavior is to inject words like \"revolutionize\", \"seamlessly\", or worse\u2014invent fake round numbers like \"boosts ROI by 300%\". It generates \"fake confidence.\" PrecisionAudit is a strict, dual-agent system designed to do the", "context": "", "use_case": "", "switched_from": "", "story": "Show HN: PrecisionAudit\u2013An AI copy auditor that actively deletes marketing fluff"}, {"id": 755612962, "item_id": 47225810, "url": "https://news.ycombinator.com/item?id=47225810", "by": "abangser", "date": "2026-03-02", "version": null, "stance": "neutral", "quote": "", "excerpt": "Hey nice site. The Ledge (https://the-ledge.ai) is a site I developed over the past year with more of a focus on bill content. Ultimately I found it useful to use all three LLM (Gemini, Claude, OpenAi) to ingest and dissect bill text en route to producing useful reports. Especially challenging for bills that run 1000+ pages.", "context": "", "use_case": "Ingesting and dissecting bill text", "switched_from": "", "story": "Show HN: Govbase \u2013 Follow a bill from source text to news bias to social posts"}, {"id": 754629744, "item_id": 47164359, "url": "https://news.ycombinator.com/item?id=47164359", "by": "qudent", "date": "2026-02-26", "version": null, "stance": "recommend", "quote": "When I use OpenAI, Openrouter etc., I can put 10 $ on my API key", "excerpt": "I think the fact that it is not possible to put hard spending caps on API keys might be ruled illegal by some EU court soon enough, at least when they sell to consumers (given the explosion of vibecoding end-users making some apps). When I use OpenAI, Openrouter etc., I can put 10 $ on my API key, and when the key leaks, someone can use these 10 $ and that's it. With Google, there is no way to do that - there are ext", "context": "", "use_case": "LLM API access", "switched_from": "", "story": "Google API Keys Weren't Secrets. But Then Gemini Changed the Rules"}, {"id": 753632384, "item_id": 47102024, "url": "https://news.ycombinator.com/item?id=47102024", "by": "keithluu", "date": "2026-02-21", "version": null, "stance": "warn", "quote": "I found that too much of a hassle and switched to other LLM providers.", "excerpt": "I believe OpenAI used Persona during the verification step that you must complete to use their SOTA models in the API. Not sure if it's still the case now. Anyway, I found that too much of a hassle and switched to other LLM providers.", "context": "", "use_case": "LLM API access", "switched_from": "", "story": "I Verified My LinkedIn Identity. Here's What I Handed Over"}, {"id": 752573889, "item_id": 47035868, "url": "https://news.ycombinator.com/item?id=47035868", "by": "Departed7405", "date": "2026-02-16", "version": null, "stance": "warn", "quote": "only have this models with OpenAI models. That's why I don't use them.", "excerpt": "I use models with OpenRouter, and only have this models with OpenAI models. That's why I don't use them.", "context": "", "use_case": "", "switched_from": "", "story": "I want to wash my car. The car wash is 50 meters away. Should I walk or drive?"}, {"id": 752475633, "item_id": 47029727, "url": "https://news.ycombinator.com/item?id=47029727", "by": "jnaina", "date": "2026-02-15", "version": null, "stance": "neutral", "quote": "use my OpenAI API and Gemini API access for sub-agent work", "excerpt": "I have Claude Max subscription for the main again tasks. Also use my OpenAI API and Gemini API access for sub-agent work. Once my Olares One is here, will also be using local LLMs on open models. https://one.olares.com/", "context": "", "use_case": "Sub-agent work", "switched_from": "", "story": "OpenClaw, OpenAI and the Future"}, {"id": 752019856, "item_id": 47001241, "url": "https://news.ycombinator.com/item?id=47001241", "by": "Doohickey-d", "date": "2026-02-13", "version": null, "stance": "recommend", "quote": "it works quite well for low-priority things", "excerpt": "Openai has a \"flex\" processing tier, which works like the normal API, but where you accept higher latency and higher error rates, in exchange for 50% off (same as batch pricing). It also supports prompt caching for further savings. For me, it works quite well for low-priority things, without the hassle of using the batch API. Usually the added latency is just a few seconds extra, so it would still work in an agent lo", "context": "", "use_case": "low-priority requests", "switched_from": "", "story": "GPT\u20115.3\u2011Codex\u2011Spark"}, {"id": 751814597, "item_id": 46988412, "url": "https://news.ycombinator.com/item?id=46988412", "by": "uokesita", "date": "2026-02-12", "version": null, "stance": "neutral", "quote": "OpenAI", "excerpt": "Location: Barcelona, Spain Remote: 100% Willing to relocate: No Technologies: ReactJS, React Native, Ruby On Rails. AI (Langchain, OpenAI, AI SDK) Tech Stack: AWS, Bedrock, Amplify, AppSync, ReactJS, NextJS, Langchain, Langgraph, OpenAI, GeminiAI, Mistral. Website: https://osledybazo.com Email: osledybazo@gmail.com", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: Who wants to be hired? (February 2026)"}, {"id": 750919712, "item_id": 46932482, "url": "https://news.ycombinator.com/item?id=46932482", "by": "ponyous", "date": "2026-02-08", "version": null, "stance": "recommend", "quote": "improves the speed of a request by about 40% according to my tests", "excerpt": "I see no mention of that, but OpenAI already has \"service tier\" API option[0] that improves the speed of a request by about 40% according to my tests. [0]: https://openai.com/api-priority-processing/", "context": "", "use_case": "Faster API requests (service tier)", "switched_from": "", "story": "Speed up responses with fast mode"}, {"id": 750824132, "item_id": 46926508, "url": "https://news.ycombinator.com/item?id=46926508", "by": "Halinani8", "date": "2026-02-07", "version": null, "stance": "neutral", "quote": "OpenAI for profile Embedding.", "excerpt": "Hi HN I\u2019ve been building NextMatchDating, a dating app centered around short, live 5-minute video conversations instead of endless swiping and texting. The idea came from noticing how much time people spend chatting without ever meeting \u2014 and how often conversations just fade out. How it works: You\u2019re matched and immediately enter a live video date The date lasts 5 minutes After 90 seconds, a \u201cNext\u201d button activates ", "context": "", "use_case": "profile embeddings", "switched_from": "", "story": "NextMatch \u2013 5-minute video speed dating to reduce ghosting"}, {"id": 746463712, "item_id": 46653982, "url": "https://news.ycombinator.com/item?id=46653982", "by": "tlarkworthy", "date": "2026-01-16", "version": null, "stance": "recommend", "quote": "OpenAI API lets you do regex structured output and it's much better than JSON for code.", "excerpt": "I use regex to force an XML schema and then use a normal XML parser to decode. XML is better for code, and for code parts in particular I enforce a cdata[[ part so there LLM is pretty free to do anything without escaping. OpenAI API lets you do regex structured output and it's much better than JSON for code.", "context": "", "use_case": "Structured output", "switched_from": "", "story": "LLM Structured Outputs Handbook"}, {"id": 745892224, "item_id": 46618264, "url": "https://news.ycombinator.com/item?id=46618264", "by": "spacecadet", "date": "2026-01-14", "version": null, "stance": "warn", "quote": "OpenAI was still the fastest, cheapest, working platform for it up until middle of last year.", "excerpt": "Its true. We have basically moved off the platforms for agentic security and host our own models now... OpenAI was still the fastest, cheapest, working platform for it up until middle of last year. Hey OpenAI, thank us later for blasting your platform with threat actor data and behavior for several years! :P", "context": "", "use_case": "", "switched_from": "", "story": null}]}