[{"data":1,"prerenderedAt":651},["ShallowReactive",2],{"project-46713106":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":45,"post_text":48,"what":49,"said":85,"maker_in_thread":187,"shape":189,"grid":194,"attempts":614,"same_week":615,"maker_shows":38,"maker_hidden":643,"reader_version":644,"card":645},"46713106","Sweep Next-Edit 1.5B","Sweep, Open-weights 1.5B model for next-edit autocomplete","title","Show HN: Sweep, Open-weights 1.5B model for next-edit autocomplete","https:\u002F\u002Fhuggingface.co\u002Fsweepai\u002Fsweep-next-edit-1.5B","huggingface.co","williamzeng0",1769037760,2026,534,153,43,4366,1,"legend","ai","code autocomplete model","library",[24,25,26,27],"next edit autocomplete","open weights code model","local code completion","fim autocomplete","curious",0.219,{"positive":31,"neutral":32,"mixed":18,"critical":33},23,69,5,98,{"performance":36,"execution":37,"privacy":39,"idea":40,"docs":42,"pricing":43},{"positive":33,"critical":18},{"positive":38,"critical":18},2,{"positive":18,"critical":18},{"positive":41,"critical":18},8,{"positive":18,"critical":18},{"positive":18,"critical":44},0,[33,46,33,44,44,44,44,44,46,18,33,46,47,47,47,47,46,47,47,46,47,33,46,46,33,46,47,47,47,47,47,33,47,47,47,46,46,33,46,38],4,3,"Hey HN, we trained and open-sourced a 1.5B model that predicts your next edits, similar to Cursor. You can download the weights here (https:\u002F\u002Fhuggingface.co\u002Fsweepai\u002Fsweep-next-edit-1.5b) or try it in our JetBrains plugin (https:\u002F\u002Fplugins.jetbrains.com\u002Fplugin\u002F26860-sweep-ai-autocomplete--coding-agent).\n\nNext-edit autocomplete differs from standard autocomplete by using your recent edits as context when predicting completions. The model is small enough to run locally while outperforming models 4x its size on both speed and accuracy.\n\nWe tested against Mercury (Inception), Zeta (Zed), and Instinc",{"problem":50,"audience":56,"pricing":60,"motivation":61,"time_to_build":65,"stack":69,"openness":82,"stage":83,"team_size":84},{"value":51,"source":52,"evidence":53},"Predicting developers' next code edits using recent edits as context, runnable locally.","maker",[54],{"id":55},"post",{"value":57,"source":52,"evidence":58},"Developers wanting local or privacy-preserving code autocomplete in editors.",[59],{"id":55},null,{"value":62,"source":52,"evidence":63},"Enable fast, privacy-preserving autocomplete for any editor.",[64],{"id":55},{"value":66,"source":52,"evidence":67},"SFT took 4 hours on 8xH100, followed by RL for 2000 steps.",[68],{"id":55},[70,78],{"value":71,"source":72,"evidence":73},"Qwen2.5-Coder base","commenter",[74],{"id":75,"by":76,"text":77},"46715461","dcreater","Based on qwen2.5-coder? seems like a \"why not\u002Fresume embellish\u002Fshow VC\" type release I guess",{"value":79,"source":52,"evidence":80},"tree-sitter parse checking in RL",[81],{"id":55},"open_source","launched","company",{"praise":86,"objections":110,"requests":125,"questions":152,"alternatives":171},[87,102],{"value":88,"source":72,"evidence":89},"Strong performance and quality for a small local model.",[90,94,98],{"id":91,"by":92,"text":93},"46717434","leonardcser","Hi, I tried the model and I am super impressed by the performance\u002Fquality. Thanks for making this open source! I am the author of this Neovim plugin for edit completions. I was…",{"id":95,"by":96,"text":97},"46717686","notsylver","I've been waiting for something like this for ages. Cursor making me pay $20\u002Fmonth when all I use from it is autocomplete was always a little annoying, especially as they changed…",{"id":99,"by":100,"text":101},"46717248","ttoinou","Wow, I can even chat about C code with that model with LM Studio on my Macbook at 200 tokens per seconds",{"value":103,"source":72,"evidence":104},"Useful open weights that could replace paid autocomplete subscriptions.",[105,109],{"id":106,"by":107,"text":108},"46716732","vanillameow","Sometimes when I use a plugin like this I get reminded just how much of a productivity nerf it is to code without an autocomplete AI. Honestly in my opinion if you write a lot of…",{"id":95,"by":96,"text":97},[111,118],{"value":112,"source":72,"evidence":113},"Suggestions are overeager and some code quality is poor for C#.",[114],{"id":115,"by":116,"text":117},"46731492","Semaphor","At least for C#, the quality of the cloud offering is rather mediocre, so I don’t expect this model to be that useful there. It’s very overeager, suggesting tons of stuff that I…",{"value":119,"source":72,"evidence":120},"Suggestions are not impressive compared to larger models.",[121],{"id":122,"by":123,"text":124},"46720568","cmrdporcupine","I've been trying my hands at implementing an emacs package for inline completions with this. I have it mostly working and performance is good enough but I haven't been blown away…",[126,141],{"value":127,"source":72,"evidence":128},"Support for VSCode and Neovim integration.",[129,133,137],{"id":130,"by":131,"text":132},"46715554","syntaxing","Wow super fun read, I love how it went into the technical details. Any way to make it work with vscode?",{"id":134,"by":135,"text":136},"46715863","dainiusse","Any easy way to try on vscode?",{"id":138,"by":139,"text":140},"46746855","dubesar55","Has somebody built any vscode extensions for this? Also is anyone serving this model?",{"value":142,"source":72,"evidence":143},"Ollama or LM Studio support.",[144,148],{"id":145,"by":146,"text":147},"46717623","woile","Hey, ollama run as suggested in hf doesn't seem to work with this model. This worked instead: ollama pull hf.co\u002Fsweepai\u002Fsweep-next-edit-1.5B",{"id":149,"by":150,"text":151},"46717616","keyle","I'm playing around with this in LMStudio (in huggingface -> use this model dropdown -> LMStudio) It's really impressive so far, so quick to respond on a mac mini M2. And it…",[153,160],{"value":154,"source":72,"evidence":155},"Is the training data available?",[156],{"id":157,"by":158,"text":159},"46717747","zoobab","Where is the training data? We can't keep calling those models \"open source\" if we have a black box and know precisely how they were made. \"Open weights\" are the new binary.",{"value":161,"source":72,"evidence":162},"Does the JetBrains plugin run offline or phone home?",[163,167],{"id":164,"by":165,"text":166},"46716112","kleiba","Very cool! I understand that the 1.5B is small enough to run locally... but does it actually in the Sweep AI Jetbrains plugin? That is, if I install the plugin, will I download…",{"id":168,"by":169,"text":170},"46719869","smusamashah","Can this be used offline in Jetbrain IntelliJ? Looking at the plugin, it looks like it requires sign in and then it uses the cloud based model instead of the local one. Can't tell.",[172,180],{"value":173,"source":72,"evidence":174},"GitHub Copilot",[175,176],{"id":106,"by":107,"text":108},{"id":177,"by":178,"text":179},"46716773","jedisct1","Really cool. But how to use it instead of Copilot in VSCode ?",{"value":181,"source":72,"evidence":182},"Continue.dev",[183],{"id":184,"by":185,"text":186},"46716519","KronisLV","I remember using Qwen 2.5 Coder for autocomplete with Continue.dev, that experience was a mess both in JetBrains IDEs, as well as Visual Studio Code. People posting stuff like…",{"comments":44,"answered":44,"key":188},[],{"commenters":190,"max_depth":191,"maker_replies":44,"first_comment_mins":192,"lifespan_hours":193},92,7,314,338,[195,197,201,205,208,211,215,218,220,223,225,227,230,233,235,237,239,241,243,246,249,251,253,256,259,262,265,268,271,273,274,277,279,281,285,289,292,294,299,302,306,310,312,316,319,322,325,327,332,335,338,341,343,346,349,351,354,356,361,364,366,368,372,376,379,382,384,387,389,394,397,401,403,405,409,412,414,417,419,421,425,428,430,432,434,438,441,444,447,449,451,453,455,459,463,466,469,471,475,477,479,481,483,485,488,492,495,497,500,502,506,508,510,513,516,519,521,523,525,527,529,534,537,539,541,544,547,549,552,554,558,562,566,568,572,574,578,582,585,589,591,593,597,601,603,606,609,612],[91,44,33,196,92,44],"project\u002Fpositive\u002Fperformance",[198,44,44,199,200,44],"46721468","commenter\u002Fneutral\u002Fnone","lasgawe",[202,44,47,203,204,44],"46722670","project\u002Fneutral\u002Fnone","treyd",[206,44,47,207,92,44],"46723010","project\u002Fneutral\u002Fexecution",[209,44,47,203,210,44],"46724196","9999gold",[212,44,33,213,214,44],"46724860","project\u002Fpositive\u002Fnone","kevinlu1248",[216,18,47,203,217,44],"46894318","zekejohn",[184,38,44,219,185,44],"other\u002Fneutral\u002Fnone",[221,38,44,219,222,44],"46717322","lostmsu",[106,47,33,224,107,44],"project\u002Fpositive\u002Fprivacy",[226,47,44,219,123,44],"46718092",[228,47,44,219,229,44],"46718934","qorrect",[231,47,44,199,232,44],"46719450","sitkack",[234,47,44,219,123,44],"46720495",[236,47,44,219,185,44],"46724726",[238,47,44,219,123,44],"46725023",[240,47,44,219,214,44],"46727575",[242,47,44,219,214,44],"46727563",[244,47,44,219,245,44],"46719118","zarzavat",[247,47,44,219,248,44],"46720562","genghisjahn",[250,47,44,219,245,44],"46722368",[252,47,44,219,214,44],"46727435",[254,47,44,219,255,44],"46724571","norir",[257,47,44,219,258,44],"46719691","wwfn",[260,47,44,219,261,44],"46722566","djfdat",[263,47,44,219,264,44],"46719773","mark_l_watson",[266,47,44,219,267,44],"46720093","pdyc",[269,47,44,219,270,44],"46722609","esafak",[272,47,44,219,255,44],"46724586",[95,46,33,196,96,44],[275,46,47,276,214,44],"46727042","maker\u002Fneutral\u002Fexecution",[278,46,47,203,135,44],"46731736",[164,33,47,280,165,44],"project\u002Fneutral\u002Fprivacy",[282,33,18,283,284,44],"46716965","project\u002Fcritical\u002Fexecution","bjarteaarmolund",[286,33,18,287,288,44],"46717489","project\u002Fcritical\u002Fprivacy","NewsaHackO",[290,33,44,199,291,44],"46720502","rkagerer",[293,33,47,203,214,44],"46727050",[295,296,33,297,298,44],"46715994",6,"project\u002Fpositive\u002Fidea","esquire_900",[300,296,47,301,214,44],"46727056","maker\u002Fneutral\u002Fnone",[303,191,33,304,305,44],"46715619","maker\u002Fpositive\u002Fnone","martianlantern",[307,191,47,308,309,44],"46721143","project\u002Fneutral\u002Fdocs","knowaveragejoe",[311,191,47,308,214,44],"46727064",[313,41,33,314,315,44],"46715675","project\u002Fpositive\u002Fdocs","_ache_",[317,41,47,203,318,44],"46718066","mromanuk",[320,41,44,219,321,44],"46718426","evanreichard",[323,41,44,219,324,44],"46741579","jmanandchanny",[326,41,47,203,214,44],"46727065",[328,329,47,330,331,44],"46715374",9,"project\u002Fneutral\u002Fidea","kamranjon",[333,329,44,219,334,44],"46718898","sheepscreek",[336,329,44,199,337,44],"46719098","aidos",[339,329,47,203,340,44],"46721001","evolving-silica",[342,329,47,330,214,44],"46735768",[122,344,47,345,123,44],10,"project\u002Fmixed\u002Fperformance",[347,344,44,199,348,44],"46724207","mijoharas",[350,344,44,199,123,44],"46724467",[352,344,44,353,214,44],"46727261","commenter\u002Fneutral\u002Fdocs",[355,344,44,199,123,44],"46727367",[357,358,33,359,360,44],"46717130",11,"maker\u002Fpositive\u002Fidea","logicallee",[157,362,18,363,158,44],12,"project\u002Fcritical\u002Fdocs",[365,362,47,308,214,44],"46727074",[168,367,47,280,169,44],13,[369,370,33,297,371,44],"46721529",14,"jrop",[373,374,47,203,375,44],"46716485",15,"magnat",[377,374,47,203,378,44],"46716823","denysvitali",[380,374,47,330,381,44],"46716914","pezgrande",[383,374,44,199,375,44],"46716962",[385,374,44,199,386,44],"46717053","KeplerBoy",[388,374,44,199,100,44],"46717253",[390,391,47,392,393,44],"46716096",16,"project\u002Fneutral\u002Fpricing","h33t-l4x0r",[395,391,44,219,396,44],"46716605","BoredPositron",[398,391,44,399,400,44],"46719512","commenter\u002Fneutral\u002Fpricing","gunalx",[145,402,47,308,146,44],17,[404,402,33,213,146,44],"46719517",[406,402,47,407,408,44],"46719625","maker\u002Fneutral\u002Fdocs","theophaniel",[410,402,47,407,411,44],"46720194","Imustaskforhelp",[413,402,47,308,411,44],"46722160",[415,402,44,353,416,44],"46722396","mika6996",[418,402,47,308,411,44],"46733746",[420,402,47,308,146,44],"46724004",[422,402,33,423,424,44],"46734093","project\u002Fpositive\u002Fexecution","ihales",[426,402,33,213,427,44],"46735112","oakesm9",[429,402,47,330,424,44],"46738889",[431,402,47,407,214,44],"46727096",[433,402,47,301,214,44],"46727143",[435,436,33,423,437,44],"46715446",18,"mgz",[439,436,47,392,440,44],"46715799","8n4vidtmkvmk",[442,436,33,443,437,44],"46718734","project\u002Fpositive\u002Fpricing",[445,436,47,203,446,44],"46723526","hdjrudni",[448,436,47,280,169,44],"46716597",[130,450,33,304,131,44],19,[149,452,33,196,150,44],20,[454,452,47,301,214,44],"46727158",[456,457,47,392,458,44],"46715549",21,"bangaladore",[460,461,47,330,462,44],"46716709",22,"andruby",[464,461,47,407,465,44],"46722442","bradfa",[467,31,33,297,468,44],"46719658","ajayarama",[470,31,47,301,214,44],"46727142",[472,473,44,219,474,44],"46717182",24,"keepamovin",[476,473,44,219,214,44],"46727129",[478,473,44,219,474,44],"46728278",[480,473,44,219,214,44],"46739776",[482,473,44,219,474,44],"46752242",[134,484,47,203,135,44],25,[486,487,47,330,267,44],"46720173",26,[489,490,47,207,491,44],"46715626",27,"sim04ful",[493,490,47,207,494,44],"46715740","NitpickLawyer",[496,490,47,207,214,44],"46727257",[115,498,18,499,116,44],28,"project\u002Fcritical\u002Fperformance",[501,498,47,301,214,44],"46735759",[503,504,33,359,505,44],"46720010",29,"deepsquirrelnet",[507,504,47,301,214,44],"46727138",[177,509,47,203,178,44],30,[511,509,47,203,512,44],"46717518","flanked-evergl",[514,509,47,308,515,44],"46717847","replete",[517,509,44,199,518,44],"46718253","BoredomIsFun",[520,509,44,199,416,44],"46722452",[522,509,44,199,518,44],"46722604",[99,524,33,196,100,44],31,[526,524,47,301,214,44],"46727168",[528,524,44,199,100,44],"46730532",[530,531,47,532,533,44],"46716623",32,"project\u002Fneutral\u002Fperformance","vichle",[535,531,47,532,536,44],"46716631","jychang",[538,531,47,532,518,44],"46718243",[540,531,47,532,400,44],"46719483",[542,531,47,532,543,44],"46716693","bodegajed",[545,531,47,532,546,44],"46716719","moffkalast",[548,531,47,532,465,44],"46722560",[550,531,44,551,546,44],"46724114","commenter\u002Fneutral\u002Fperformance",[553,531,47,532,214,44],"46727199",[555,556,47,330,557,44],"46715699",33,"whimsicalism",[559,560,33,359,561,44],"46715694",34,"_boffin_",[563,564,47,330,565,44],"46737125",35,"k9294",[138,567,47,203,139,44],36,[569,570,47,330,571,44],"46718803",37,"bberenberg",[573,570,33,297,214,44],"46727268",[575,576,47,203,577,44],"46716816",38,"ragchronos",[579,580,44,219,581,44],"46716205",39,"moelf",[583,580,44,219,584,44],"46716481","dajonker",[586,587,47,203,588,44],"46715649",40,"rationably",[590,587,33,196,315,44],"46715698",[592,587,47,203,214,44],"46727279",[594,595,33,359,596,44],"46717627",41,"_mugencode",[598,599,47,203,600,44],"46717187",42,"rw_panic0_0",[602,599,47,203,214,44],"46727353",[604,16,47,203,605,44],"46715494","ing33k",[75,607,18,608,76,44],44,"project\u002Fcritical\u002Fidea",[610,607,44,199,611,44],"46715512","dang",[613,607,47,532,214,44],"46727316",[],[616,621,625,630,635,640],{"id":617,"name":618,"points":619,"category":620},"46706528","ChartGPU",670,"data",{"id":622,"name":623,"points":192,"category":624},"46673809","Calquio","finance_crypto",{"id":626,"name":627,"points":628,"category":629},"46696283","Coi",225,"languages",{"id":631,"name":632,"points":633,"category":634},"46709543","Rails UI",208,"dev_tools",{"id":636,"name":637,"points":638,"category":639},"46671952","Dock",177,"productivity",{"id":641,"name":642,"points":638,"category":639},"46675231","Pdfwithlove",false,"v2",{"id":646,"name":5,"line":6,"ai":643,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":647,"top":18,"kw":648,"maker":11,"domain":10,"tier":19,"attempts":44,"strip":649,"url":9,"mix":650,"praise":88,"objection":112,"sub":21},46713106,0.22,[24,25,26,27],"5450000041543333433435445433333533344542",[31,32,18,33],1794036690944]