[{"data":1,"prerenderedAt":161},["ShallowReactive",2],{"project-41118609":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":28,"sentiment":29,"stance_mix":30,"scored_comments":34,"aspects":35,"strip":39,"post_text":41,"what":42,"said":67,"maker_in_thread":108,"shape":111,"grid":114,"attempts":125,"same_week":126,"maker_shows":15,"maker_hidden":154,"reader_version":155,"card":156},"41118609","Neurooo","DeepL clone using LLMs for translations","title","Show HN: Neurooo – DeepL clone using LLMs for translations","https:\u002F\u002Fneurooo.com\u002F","neurooo.com","MichaelHoste",1722428723,2024,18,4,1990,3515,57,"shown","languages","translation tool","web_app",[24,25,26,27],"llm translation","deepl alternative","translation api","gpt translation","quiet",null,{"positive":31,"neutral":32,"mixed":32,"critical":33},2,0,1,3,{"execution":36,"idea":37,"performance":38},{"positive":32,"critical":33},{"positive":33,"critical":32},{"positive":33,"critical":32},[33,32,40,40],5,"Hi HN,\n\nWe built Neurooo.com, an automatic translation tool like DeepL or Google Translate, but using LLMs with OpenAI and Mistral.\n\nWe noticed that LLMs were quite good at translating informal or contextual text, and also way cheaper than DeepL or Google Translate when you factor in the cost per character.\n\nSo we decided to try and build a similar translation interface to see if we can match the leaders in term of ease of use and quality.\n\nWe're quite happy about it and people around us seem to think that we're even better in many cases (abbreviations, informal emails, business translations, ",{"problem":43,"audience":29,"pricing":29,"motivation":49,"time_to_build":29,"stack":53,"openness":65,"stage":66,"team_size":65},{"value":44,"source":45,"evidence":46},"Translating informal or contextual text cheaply with LLMs instead of DeepL or Google Translate","maker",[47],{"id":48},"post",{"value":50,"source":45,"evidence":51},"LLMs translate informal text well and cost less per character than DeepL or Google Translate",[52],{"id":48},[54,61],{"value":55,"source":45,"evidence":56},"OpenAI models",[57,58],{"id":48},{"id":59,"by":11,"text":60},"41168934","Thanks for your feedback! Yes, we're using ChatGPT 3.5 for now and we will certainly switch to GPT-4o mini in the near future. It already works with ChatGPT 4 but we noticed that…",{"value":62,"source":45,"evidence":63},"Mistral models",[64],{"id":48},"unknown","launched",{"praise":68,"objections":81,"requests":89,"questions":90,"alternatives":95},[69,73],{"value":70,"source":45,"evidence":71},"Translations of informal and business text seem better than DeepL in some cases",[72],{"id":48},{"value":74,"source":75,"evidence":76},"Fast and stable API in a medical app for East European translation","commenter",[77],{"id":78,"by":79,"text":80},"41124072","frbayart","We use it in our medical app to translate it into the East European language, and the API is very fast and stable! No need to search for an alternative",[82],{"value":83,"source":75,"evidence":84},"Japanese translation was inaccurate on a family-relationship sentence",[85],{"id":86,"by":87,"text":88},"41137583","gerdemb","Very cool! I've transitioned from using Google Translate and DeepL to ChatGPT for translations. In general I think the ChatGPT translations are better, but the default UI for is…",[],[91],{"value":92,"source":75,"evidence":93},"Which ChatGPT model is used, 3.5 or 4",[94],{"id":86,"by":87,"text":88},[96,100,104],{"value":97,"source":75,"evidence":98},"Google Translate",[99],{"id":86,"by":87,"text":88},{"value":101,"source":75,"evidence":102},"DeepL",[103],{"id":86,"by":87,"text":88},{"value":105,"source":75,"evidence":106},"ChatGPT",[107],{"id":86,"by":87,"text":88},{"comments":33,"answered":33,"key":109},[110],{"id":59,"text":60},{"commenters":15,"max_depth":31,"maker_replies":33,"first_comment_mins":112,"lifespan_hours":113},208,140,[115,117,119,123],[86,32,33,116,87,32],"project\u002Fcritical\u002Fexecution",[59,32,32,118,11,33],"maker\u002Fneutral\u002Fexecution",[120,33,40,121,122,32],"41120289","project\u002Fpositive\u002Fidea","m4thias",[78,31,40,124,79,32],"project\u002Fpositive\u002Fperformance",[],[127,132,136,140,145,150],{"id":128,"name":129,"points":130,"category":131},"41068719","Haystack",546,"dev_tools",{"id":133,"name":134,"points":135,"category":131},"41070428","Wat",393,{"id":137,"name":138,"points":139,"category":131},"41088273","Semantic Grep",356,{"id":141,"name":142,"points":143,"category":144},"41086620","Opslane",319,"infrastructure",{"id":146,"name":147,"points":148,"category":149},"41105130","embedding.io",305,"ai",{"id":151,"name":152,"points":153,"category":131},"41080189","gg",205,false,"v2",{"id":157,"name":5,"line":6,"ai":154,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":29,"top":18,"kw":158,"maker":11,"domain":10,"tier":19,"attempts":32,"strip":159,"url":9,"mix":160,"praise":70,"objection":83,"sub":21},41118609,[24,25,26,27],"1055",[31,32,32,33],1794036658638]