{"slug": "gemini-2-5-flash", "name": "Gemini 2.5 Flash", "category": "ai", "category_name": "AI models and assistants", "description": "Google's fast, low-cost multimodal large language model in the Gemini 2.5 family.", "recommenders": 9, "commenters": 11, "warners": 0, "score": 8.456, "rising": -0.29, "first": "2025-11-20", "latest": "2026-10-29", "why": "HN commenters like it for low latency and low cost, saying nothing in its price range matches its all-around performance.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 9, "neutral": 2, "compare": 0, "warn": 0}, "spark": [1, 2, 1, 1, 1, 0, 0, 0, 1, 0, 0, 2], "listed": true, "loved": ["Speed", "Price-performance"], "caveat": "Deprecation", "short_quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "icon": true, "domain": null, "github_repo": null, "official_url": "https://deepmind.google/models/gemini/flash/", "pricing": "paid", "aliases": ["Gemini 2.5 Flash"], "verdict": [{"text": "HN commenters like it for low latency and low cost,", "cites": [{"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}, {"id": 758389312, "item_id": 47399332, "url": "https://news.ycombinator.com/item?id=47399332", "by": "hamdingers", "date": "2026-03-16", "stance": "recommend", "quote": "I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant", "excerpt": "If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra", "context": "", "use_case": "Home Assistant voice assistant", "switched_from": "", "story": "My Journey to a reliable and enjoyable locally hosted voice assistant"}, {"id": 755205153, "item_id": 47200322, "url": "https://news.ycombinator.com/item?id=47200322", "by": "aggeeinn", "date": "2026-02-28", "stance": "recommend", "quote": "I used Gemini specifically because the inference speed is incredible", "excerpt": "Hey HN, OP here. I\u2019ve been frustrated by how difficult it is to track actual, kinetic events during modern conflicts without getting buried in algorithmic noise, propaganda, or Telegram spam. I wanted to build a purely automated, objective pipeline to map verifiable events as they happen. I built IranWarLive as a serverless OSINT dashboard. Here is how the stack works under the hood: The Architecture: The Edge: A Clo", "context": "", "use_case": "parsing news into JSON", "switched_from": "", "story": "Show HN: IranWarLive \u2013 Automated, serverless OSINT mapping engine"}, {"id": 740934608, "item_id": 46308413, "url": "https://news.ycombinator.com/item?id=46308413", "by": "gorbot", "date": "2025-12-17", "stance": "recommend", "quote": "the pro was not noticeably more accurate, but significantly slower, so I used flash way more", "excerpt": "Ive been using 2.5 pro or flash a ton at work and the pro was not noticeably more accurate, but significantly slower, so I used flash way more. This is super exciting", "context": "", "use_case": "work tasks", "switched_from": "", "story": "Gemini 3 Flash: Frontier intelligence built for speed"}]}, {"text": " saying nothing in its price range matches its all-around performance.", "cites": [{"id": 781857088, "item_id": 48866068, "url": "https://news.ycombinator.com/item?id=48866068", "by": "zuzululu", "date": "2026-07-10", "stance": "recommend", "quote": "theres nothing in its price range that provides the same all around perf", "excerpt": "sucks we use 2.5 flash/lite in our company it handles millions of requests a day theres nothing in its price range that provides the same all around perf as noted, gemini 3 flash is expensive really not liking google these days they are not hungry anymore", "context": "company using it at scale", "use_case": "millions of requests a day", "switched_from": "", "story": "Don't discontinue Gemini 2.5 Flash"}, {"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}, {"id": 804621440, "item_id": 50288840, "url": "https://news.ycombinator.com/item?id=50288840", "by": "pcwelder", "date": "2026-10-29", "stance": "recommend", "quote": "Gemini 2.5 flash and 3 flash were highly underrated models that performed way better than the peers", "excerpt": "Gemini 2.5 flash and 3 flash were highly underrated models that performed way better than the peers and we used the hell out of them. Since then none of the Google models have competed well with the others at their price points. Long context performance drop, instruction following issues, unnecessarily complex and sloppish outputs. Argon is no different, it feels just an iteration better than 3.8 flash. I don't trust", "context": "", "use_case": "production LLM workloads", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}]}, {"text": " The main caveat is that Google is deprecating it,", "cites": [{"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}]}, {"text": " and one commenter finds the successor, Gemini 3 Flash, too expensive.", "cites": [{"id": 781857088, "item_id": 48866068, "url": "https://news.ycombinator.com/item?id=48866068", "by": "zuzululu", "date": "2026-07-10", "stance": "recommend", "quote": "theres nothing in its price range that provides the same all around perf", "excerpt": "sucks we use 2.5 flash/lite in our company it handles millions of requests a day theres nothing in its price range that provides the same all around perf as noted, gemini 3 flash is expensive really not liking google these days they are not hungry anymore", "context": "company using it at scale", "use_case": "millions of requests a day", "switched_from": "", "story": "Don't discontinue Gemini 2.5 Flash"}]}], "praise": [{"label": "Speed", "summary": "Commenters value its low latency and fast inference.", "quotes": [{"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}, {"id": 755205153, "item_id": 47200322, "url": "https://news.ycombinator.com/item?id=47200322", "by": "aggeeinn", "date": "2026-02-28", "stance": "recommend", "quote": "I used Gemini specifically because the inference speed is incredible", "excerpt": "Hey HN, OP here. I\u2019ve been frustrated by how difficult it is to track actual, kinetic events during modern conflicts without getting buried in algorithmic noise, propaganda, or Telegram spam. I wanted to build a purely automated, objective pipeline to map verifiable events as they happen. I built IranWarLive as a serverless OSINT dashboard. Here is how the stack works under the hood: The Architecture: The Edge: A Clo", "context": "", "use_case": "parsing news into JSON", "switched_from": "", "story": "Show HN: IranWarLive \u2013 Automated, serverless OSINT mapping engine"}, {"id": 758389312, "item_id": 47399332, "url": "https://news.ycombinator.com/item?id=47399332", "by": "hamdingers", "date": "2026-03-16", "stance": "recommend", "quote": "I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant", "excerpt": "If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra", "context": "", "use_case": "Home Assistant voice assistant", "switched_from": "", "story": "My Journey to a reliable and enjoyable locally hosted voice assistant"}]}, {"label": "Price-performance", "summary": "Users say nothing at its price matches its overall performance.", "quotes": [{"id": 781857088, "item_id": 48866068, "url": "https://news.ycombinator.com/item?id=48866068", "by": "zuzululu", "date": "2026-07-10", "stance": "recommend", "quote": "theres nothing in its price range that provides the same all around perf", "excerpt": "sucks we use 2.5 flash/lite in our company it handles millions of requests a day theres nothing in its price range that provides the same all around perf as noted, gemini 3 flash is expensive really not liking google these days they are not hungry anymore", "context": "company using it at scale", "use_case": "millions of requests a day", "switched_from": "", "story": "Don't discontinue Gemini 2.5 Flash"}, {"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}, {"id": 804621440, "item_id": 50288840, "url": "https://news.ycombinator.com/item?id=50288840", "by": "pcwelder", "date": "2026-10-29", "stance": "recommend", "quote": "Gemini 2.5 flash and 3 flash were highly underrated models that performed way better than the peers", "excerpt": "Gemini 2.5 flash and 3 flash were highly underrated models that performed way better than the peers and we used the hell out of them. Since then none of the Google models have competed well with the others at their price points. Long context performance drop, instruction following issues, unnecessarily complex and sloppish outputs. Argon is no different, it feels just an iteration better than 3.8 flash. I don't trust", "context": "", "use_case": "production LLM workloads", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}]}, {"label": "Beats Pro models", "summary": "Some found it as accurate as or better than Gemini Pro models while being faster.", "quotes": [{"id": 740934608, "item_id": 46308413, "url": "https://news.ycombinator.com/item?id=46308413", "by": "gorbot", "date": "2025-12-17", "stance": "recommend", "quote": "the pro was not noticeably more accurate, but significantly slower, so I used flash way more", "excerpt": "Ive been using 2.5 pro or flash a ton at work and the pro was not noticeably more accurate, but significantly slower, so I used flash way more. This is super exciting", "context": "", "use_case": "work tasks", "switched_from": "", "story": "Gemini 3 Flash: Frontier intelligence built for speed"}, {"id": 735963072, "item_id": 45997692, "url": "https://news.ycombinator.com/item?id=45997692", "by": "Nemi", "date": "2025-11-20", "stance": "recommend", "quote": "2.5 flash works really well but 2.5 pro or 3.0 pro do a very poor job", "excerpt": "I feel like I am going crazy or missed something simple but when I use the Gemini app and I ask it to edit a photo that I upload, 2.5 flash works really well but 2.5 pro or 3.0 pro do a very poor job. I uploaded an image of me and asked it to make me bald and flash did a great job of just changing me in the photo but 3.0 pro took me out of the photo completely and just created a headshot of a bald man that only sort ", "context": "", "use_case": "editing uploaded photos", "switched_from": "", "story": "Nano Banana Pro"}]}, {"label": "Image tasks", "summary": "It is recommended for image segmentation and praised for photo editing.", "quotes": [{"id": 743491729, "item_id": 46468233, "url": "https://news.ycombinator.com/item?id=46468233", "by": "simonw", "date": "2026-01-02", "stance": "recommend", "quote": "we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5", "excerpt": "It's not supported by Gemini 3: https://ai.google.dev/gemini-api/docs/gemini-3#migrating_from_gemini_25 > Image segmentation: Image segmentation capabilities (returning pixel-level masks for objects) are not supported in Gemini 3 Pro or Gemini 3 Flash. For workloads requiring native image segmentation, we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5.", "context": "", "use_case": "image segmentation", "switched_from": "", "story": "OfferGridAI \u2013 side-by-side comparison of real estate offers from PDFs"}, {"id": 735963072, "item_id": 45997692, "url": "https://news.ycombinator.com/item?id=45997692", "by": "Nemi", "date": "2025-11-20", "stance": "recommend", "quote": "2.5 flash works really well but 2.5 pro or 3.0 pro do a very poor job", "excerpt": "I feel like I am going crazy or missed something simple but when I use the Gemini app and I ask it to edit a photo that I upload, 2.5 flash works really well but 2.5 pro or 3.0 pro do a very poor job. I uploaded an image of me and asked it to make me bald and flash did a great job of just changing me in the photo but 3.0 pro took me out of the photo completely and just created a headshot of a bald man that only sort ", "context": "", "use_case": "editing uploaded photos", "switched_from": "", "story": "Nano Banana Pro"}]}, {"label": "Voice use", "summary": "It is used for voice-to-text transcription and as a voice home assistant.", "quotes": [{"id": 740354659, "item_id": 46272166, "url": "https://news.ycombinator.com/item?id=46272166", "by": "k9294", "date": "2025-12-15", "stance": "recommend", "quote": "get solid voice-to-text using Gemini 2.5 Flash", "excerpt": "Working on https://ottex.ai/ - BYOK alternative to Wispr Flow and Raycast AI shortcuts. I love global voice-to-text transcription (especially when working with Claude Code or Cursor) and simple AI shortcuts like \"Fix Grammar\" and \"Translate to {Language}\". I realized I was spending around \u20ac35/mo (\u20ac420 a year) on two apps for AI features that cost just pennies to run. So I built Ottex - a native macOS app with a tiny ", "context": "", "use_case": "voice transcription", "switched_from": "", "story": "Ask HN: What Are You Working On? (December 2025)"}, {"id": 758389312, "item_id": 47399332, "url": "https://news.ycombinator.com/item?id=47399332", "by": "hamdingers", "date": "2026-03-16", "stance": "recommend", "quote": "I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant", "excerpt": "If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra", "context": "", "use_case": "Home Assistant voice assistant", "switched_from": "", "story": "My Journey to a reliable and enjoyable locally hosted voice assistant"}]}, {"label": "Handles scale", "summary": "One company runs millions of requests a day through it.", "quotes": [{"id": 781857088, "item_id": 48866068, "url": "https://news.ycombinator.com/item?id=48866068", "by": "zuzululu", "date": "2026-07-10", "stance": "recommend", "quote": "theres nothing in its price range that provides the same all around perf", "excerpt": "sucks we use 2.5 flash/lite in our company it handles millions of requests a day theres nothing in its price range that provides the same all around perf as noted, gemini 3 flash is expensive really not liking google these days they are not hungry anymore", "context": "company using it at scale", "use_case": "millions of requests a day", "switched_from": "", "story": "Don't discontinue Gemini 2.5 Flash"}]}], "caveats": [{"label": "Deprecation", "summary": "A commenter reports Google is removing the model.", "quotes": [{"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}]}, {"label": "Costly successor", "summary": "One commenter says the replacement, Gemini 3 Flash, is expensive.", "quotes": [{"id": 781857088, "item_id": 48866068, "url": "https://news.ycombinator.com/item?id=48866068", "by": "zuzululu", "date": "2026-07-10", "stance": "recommend", "quote": "theres nothing in its price range that provides the same all around perf", "excerpt": "sucks we use 2.5 flash/lite in our company it handles millions of requests a day theres nothing in its price range that provides the same all around perf as noted, gemini 3 flash is expensive really not liking google these days they are not hungry anymore", "context": "company using it at scale", "use_case": "millions of requests a day", "switched_from": "", "story": "Don't discontinue Gemini 2.5 Flash"}]}, {"label": "Privacy", "summary": "As a cloud model, one commenter suggests it suits those less concerned about privacy.", "quotes": [{"id": 758389312, "item_id": 47399332, "url": "https://news.ycombinator.com/item?id=47399332", "by": "hamdingers", "date": "2026-03-16", "stance": "recommend", "quote": "I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant", "excerpt": "If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra", "context": "", "use_case": "Home Assistant voice assistant", "switched_from": "", "story": "My Journey to a reliable and enjoyable locally hosted voice assistant"}]}, {"label": "Segmentation setup", "summary": "Google's docs recommend turning thinking off for image segmentation.", "quotes": [{"id": 743491729, "item_id": 46468233, "url": "https://news.ycombinator.com/item?id=46468233", "by": "simonw", "date": "2026-01-02", "stance": "recommend", "quote": "we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5", "excerpt": "It's not supported by Gemini 3: https://ai.google.dev/gemini-api/docs/gemini-3#migrating_from_gemini_25 > Image segmentation: Image segmentation capabilities (returning pixel-level masks for objects) are not supported in Gemini 3 Pro or Gemini 3 Flash. For workloads requiring native image segmentation, we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5.", "context": "", "use_case": "image segmentation", "switched_from": "", "story": "OfferGridAI \u2013 side-by-side comparison of real estate offers from PDFs"}]}], "who": [{"by": "skinnychef", "context": "not a dev", "url": "https://news.ycombinator.com/item?id=49761819", "date": "2026-09-18"}, {"by": "zuzululu", "context": "company using it at scale", "url": "https://news.ycombinator.com/item?id=48866068", "date": "2026-07-10"}], "tags": ["for high-volume APIs", "for voice assistants", "for image segmentation", "for home automation", "for terminal workflows"], "use_cases": ["production llm workloads", "low-latency llm tasks", "terminal ai queries", "millions of requests a day", "elderly monitoring app analysis", "home assistant voice assistant"], "compared": [{"slug": "gemma-4-26b-a4b", "name": "Gemma 4 26B A4B", "co": 1, "switched_from": 0, "switched_to": 1}, {"slug": "gemini-flash", "name": "Gemini Flash", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "gemini-pro", "name": "Gemini Pro", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=50288019", "title": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49933476", "title": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49740105", "title": "Show HN: Share your AI Setup, Learn from others", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48864507", "title": "Don't discontinue Gemini 2.5 Flash", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47791832", "title": "Ask HN: How are you using LLMs in production?", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47398534", "title": "My Journey to a reliable and enjoyable locally hosted voice assistant", "mentions": 1}], "sentiment_by_month": {"2025-11": {"recommend": 1}, "2025-12": {"recommend": 2}, "2026-01": {"recommend": 1}, "2026-02": {"recommend": 1}, "2026-03": {"recommend": 1}, "2026-04": {"neutral": 1}, "2026-05": {}, "2026-06": {}, "2026-07": {"recommend": 1}, "2026-08": {}, "2026-09": {"neutral": 1}, "2026-10": {"recommend": 2}}, "quotes": [{"id": 804621440, "item_id": 50288840, "url": "https://news.ycombinator.com/item?id=50288840", "by": "pcwelder", "date": "2026-10-29", "stance": "recommend", "quote": "Gemini 2.5 flash and 3 flash were highly underrated models that performed way better than the peers", "excerpt": "Gemini 2.5 flash and 3 flash were highly underrated models that performed way better than the peers and we used the hell out of them. Since then none of the Google models have competed well with the others at their price points. Long context performance drop, instruction following issues, unnecessarily complex and sloppish outputs. Argon is no different, it feels just an iteration better than 3.8 flash. I don't trust", "context": "", "use_case": "production LLM workloads", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}, {"id": 799512336, "item_id": 49969521, "url": "https://news.ycombinator.com/item?id=49969521", "by": "pugio", "date": "2026-10-05", "stance": "recommend", "quote": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost.", "excerpt": "I use Gemini-2.5-flash extensively, mostly for its low latency, and haven't found anything which works better at that speed/cost. Which is why I'm really bummed that Google is deprecaty/removing it next week. Have you found anything that works as a suitable replacement?", "context": "", "use_case": "low-latency LLM tasks", "switched_from": "", "story": "Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers"}, {"id": 796189106, "item_id": 49761819, "url": "https://news.ycombinator.com/item?id=49761819", "by": "skinnychef", "date": "2026-09-18", "stance": "neutral", "quote": "fabric --model gemini-2.5-flash --pattern ask_ai", "excerpt": "Well, not a dev here, so my primitive setup is to have an alias in my .bashrc that launches firefox with the prompt added to the Google AI mode, so all I have to do is Super+T (terminal), type 'ai' followed by my query. For laughs, this also allows to bake in response styles: query=\"$*\" base=\"Respond like Dr. House, do not hold back the profanities; \" I also have an even shorter alias for '?' using fabric that answer", "context": "not a dev", "use_case": "terminal AI queries", "switched_from": "", "story": "Show HN: Share your AI Setup, Learn from others"}, {"id": 781857088, "item_id": 48866068, "url": "https://news.ycombinator.com/item?id=48866068", "by": "zuzululu", "date": "2026-07-10", "stance": "recommend", "quote": "theres nothing in its price range that provides the same all around perf", "excerpt": "sucks we use 2.5 flash/lite in our company it handles millions of requests a day theres nothing in its price range that provides the same all around perf as noted, gemini 3 flash is expensive really not liking google these days they are not hungry anymore", "context": "company using it at scale", "use_case": "millions of requests a day", "switched_from": "", "story": "Don't discontinue Gemini 2.5 Flash"}, {"id": 764784048, "item_id": 47799003, "url": "https://news.ycombinator.com/item?id=47799003", "by": "sminchev", "date": "2026-04-16", "stance": "neutral", "quote": "Used LLMs: Gemini 2.5 Flash. ChatGPT 5", "excerpt": "I have done integration it two applications: (1) I use it to analyze behavior, sleep and still patterns, and detect frequently visited locations, based on recorded and provided data. It is an monitoring app for elderly people; (2) suggestions, parsing files, as database. The first two are clear. The third, instead of supporting and maintaining a database, I call the AI to give me the information; Used LLMs: Gemini 2.", "context": "", "use_case": "elderly monitoring app analysis", "switched_from": "", "story": "Ask HN: How are you using LLMs in production?"}, {"id": 758389312, "item_id": 47399332, "url": "https://news.ycombinator.com/item?id=47399332", "by": "hamdingers", "date": "2026-03-16", "stance": "recommend", "quote": "I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant", "excerpt": "If you're less concerned about privacy, I use Gemini 2.5 Flash for this and it's exceptionally good and fast as a HA assistant while being much cheaper than the electricity that would be needed to keep a 3090 awake. The thing that kills this for me (and they even mentioned it) is wake word detection. I have both the HA voice preview and FPH Satellite1 devices, plus have experimented with a few other options like a Ra", "context": "", "use_case": "Home Assistant voice assistant", "switched_from": "", "story": "My Journey to a reliable and enjoyable locally hosted voice assistant"}, {"id": 755205153, "item_id": 47200322, "url": "https://news.ycombinator.com/item?id=47200322", "by": "aggeeinn", "date": "2026-02-28", "stance": "recommend", "quote": "I used Gemini specifically because the inference speed is incredible", "excerpt": "Hey HN, OP here. I\u2019ve been frustrated by how difficult it is to track actual, kinetic events during modern conflicts without getting buried in algorithmic noise, propaganda, or Telegram spam. I wanted to build a purely automated, objective pipeline to map verifiable events as they happen. I built IranWarLive as a serverless OSINT dashboard. Here is how the stack works under the hood: The Architecture: The Edge: A Clo", "context": "", "use_case": "parsing news into JSON", "switched_from": "", "story": "Show HN: IranWarLive \u2013 Automated, serverless OSINT mapping engine"}, {"id": 743491729, "item_id": 46468233, "url": "https://news.ycombinator.com/item?id=46468233", "by": "simonw", "date": "2026-01-02", "stance": "recommend", "quote": "we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5", "excerpt": "It's not supported by Gemini 3: https://ai.google.dev/gemini-api/docs/gemini-3#migrating_from_gemini_25 > Image segmentation: Image segmentation capabilities (returning pixel-level masks for objects) are not supported in Gemini 3 Pro or Gemini 3 Flash. For workloads requiring native image segmentation, we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5.", "context": "", "use_case": "image segmentation", "switched_from": "", "story": "OfferGridAI \u2013 side-by-side comparison of real estate offers from PDFs"}, {"id": 740934608, "item_id": 46308413, "url": "https://news.ycombinator.com/item?id=46308413", "by": "gorbot", "date": "2025-12-17", "stance": "recommend", "quote": "the pro was not noticeably more accurate, but significantly slower, so I used flash way more", "excerpt": "Ive been using 2.5 pro or flash a ton at work and the pro was not noticeably more accurate, but significantly slower, so I used flash way more. This is super exciting", "context": "", "use_case": "work tasks", "switched_from": "", "story": "Gemini 3 Flash: Frontier intelligence built for speed"}, {"id": 740354659, "item_id": 46272166, "url": "https://news.ycombinator.com/item?id=46272166", "by": "k9294", "date": "2025-12-15", "stance": "recommend", "quote": "get solid voice-to-text using Gemini 2.5 Flash", "excerpt": "Working on https://ottex.ai/ - BYOK alternative to Wispr Flow and Raycast AI shortcuts. I love global voice-to-text transcription (especially when working with Claude Code or Cursor) and simple AI shortcuts like \"Fix Grammar\" and \"Translate to {Language}\". I realized I was spending around \u20ac35/mo (\u20ac420 a year) on two apps for AI features that cost just pennies to run. So I built Ottex - a native macOS app with a tiny ", "context": "", "use_case": "voice transcription", "switched_from": "", "story": "Ask HN: What Are You Working On? (December 2025)"}, {"id": 735963072, "item_id": 45997692, "url": "https://news.ycombinator.com/item?id=45997692", "by": "Nemi", "date": "2025-11-20", "stance": "recommend", "quote": "2.5 flash works really well but 2.5 pro or 3.0 pro do a very poor job", "excerpt": "I feel like I am going crazy or missed something simple but when I use the Gemini app and I ask it to edit a photo that I upload, 2.5 flash works really well but 2.5 pro or 3.0 pro do a very poor job. I uploaded an image of me and asked it to make me bald and flash did a great job of just changing me in the photo but 3.0 pro took me out of the photo completely and just created a headshot of a bald man that only sort ", "context": "", "use_case": "editing uploaded photos", "switched_from": "", "story": "Nano Banana Pro"}]}