[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"navigation":3,"url-settings":86,"updates-\u002Fupdates\u002Fnano-models":663,"updates-related-\u002Fupdates\u002Fnano-models":1716,"i-material-symbols:arrow-back-rounded":1717},{"id":4,"extension":5,"footer":6,"header":73,"meta":83,"stem":84,"__hash__":85},"navigation\u002Fdata\u002Fshared\u002Fnavigation.yml","yml",{"brand":7,"columns":13,"legal":63},{"name":8,"tagline":9,"downloadCta":10},"Pieces","The memory layer for modern work.",{"label":11,"href":12},"All downloads","\u002Fdownloads",[14,26,39,51],{"title":15,"links":16},"Product",[17,20,23],{"label":18,"href":19},"Pieces Desktop","\u002F",{"label":21,"href":22},"Pieces Enterprise","\u002Fenterprise",{"label":24,"href":25},"Pieces MCP","\u002Fmcp",{"title":27,"links":28},"Resources",[29,33,36],{"label":30,"href":31,"external":32},"Documentation","url:docs.home",true,{"label":34,"href":35},"Blog","\u002Fblog",{"label":37,"href":38,"external":32},"GitHub","url:github.org",{"title":40,"links":41},"Community",[42,45,48],{"label":43,"href":44,"external":32},"Discord","url:social.discord",{"label":46,"href":47,"external":32},"X \u002F Twitter","url:social.x",{"label":49,"href":50,"external":32},"LinkedIn","url:social.linkedin",{"title":52,"links":53},"Company",[54,57,60],{"label":55,"href":56},"About","\u002Fabout",{"label":58,"href":59},"Updates","\u002Fupdates",{"label":61,"href":62},"Contact","\u002Fcontact",[64,67,70],{"label":65,"href":66},"Privacy Policy","\u002Flegal\u002Fprivacy",{"label":68,"href":69},"Refund Policy","\u002Flegal\u002Frefund",{"label":71,"href":72},"Terms of Service","\u002Flegal\u002Fterms",{"links":74,"signIn":75,"contact":78,"cta":80},[],{"label":76,"href":77},"Sign in","url:portal.home",{"label":79,"href":62},"Get in touch",{"label":81,"href":82},"Download","url:routes.downloads",{},"data\u002Fshared\u002Fnavigation","VdirGIa8cUm6xLprbBeTKcw914JfyJjWfF-9hSVOX5Y",{"id":87,"extension":5,"links":88,"meta":660,"stem":661,"__hash__":662},"urlSettings\u002Fdata\u002Fshared\u002Furls.yml",[89,93,97,101,105,109,113,117,121,125,129,133,137,141,145,149,153,157,161,165,169,173,177,181,185,189,192,196,200,204,208,212,216,220,224,228,232,236,240,244,248,252,256,260,264,268,272,276,280,284,288,292,295,299,303,306,310,314,318,322,326,330,334,338,342,346,350,354,358,362,366,370,374,378,382,386,390,394,398,402,406,410,414,418,422,426,430,434,438,442,446,450,453,457,461,465,469,473,477,480,483,486,489,493,497,500,504,508,512,516,520,524,528,532,536,540,544,548,551,555,559,563,567,571,574,577,580,584,587,591,595,599,603,606,610,614,617,620,623,627,631,635,638,641,644,648,652,656],{"key":90,"label":91,"href":92},"downloads.desktop","Desktop download page","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Fdownload",{"key":94,"label":95,"href":96},"downloads.macOS.dmgArm64","macOS DMG Apple Silicon","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_for_x\u002Fdmg-arm64\u002Fdownload",{"key":98,"label":99,"href":100},"downloads.macOS.dmgIntel","macOS DMG Intel","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_for_x\u002Fdmg\u002Fdownload",{"key":102,"label":103,"href":104},"downloads.macOS.pkg","macOS PKG","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fmacos_packaging\u002Fpkg\u002Fdownload",{"key":106,"label":107,"href":108},"downloads.macOS.universal","Desktop macOS Universal","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_for_x\u002Fmacos-universal\u002Fdownload",{"key":110,"label":111,"href":112},"downloads.macOS.pkgArm64","Desktop macOS PKG Apple Silicon","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fmacos_packaging\u002Fpkg-pfd-arm64\u002Fdownload",{"key":114,"label":115,"href":116},"downloads.macOS.pkgIntel","Desktop macOS PKG Intel","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fmacos_packaging\u002Fpkg-pfd\u002Fdownload",{"key":118,"label":119,"href":120},"downloads.windows.appinstaller","Windows App Installer","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fappinstaller\u002Fpieces_for_x.appinstaller",{"key":122,"label":123,"href":124},"downloads.windows.exe","Windows EXE","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_for_x\u002Fwindows-exe\u002Fdownload",{"key":126,"label":127,"href":128},"downloads.windows.msix","Windows MSIX","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_for_x\u002Fwindows-msix\u002Fdownload",{"key":130,"label":131,"href":132},"downloads.windows.suiteManager","Windows Suite Manager","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_suite_windows\u002Fappinstaller\u002Fdownload",{"key":134,"label":135,"href":136},"downloads.linux.flatpakRepo","Linux Flatpak repository","https:\u002F\u002Fbuilds.pieces.app\u002Fpieces-flatpak-repo\u002Fpieces-flatpak.flatpakrepo",{"key":138,"label":139,"href":140},"downloads.linux.snapDesktop","Linux Snap Desktop","https:\u002F\u002Fsnapcraft.io\u002Fpieces-for-developers",{"key":142,"label":143,"href":144},"downloads.linux.snapPiecesOS","Linux Snap PiecesOS","https:\u002F\u002Fsnapcraft.io\u002Fpieces-os",{"key":146,"label":147,"href":148},"downloads.linux.snap","Desktop Linux Snap package","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fpieces_for_x\u002Fsnap\u002Fdownload",{"key":150,"label":151,"href":152},"downloads.piecesOS.macOS.dmgArm64","PiecesOS macOS DMG Apple Silicon","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fos_server\u002Fdmg-arm64\u002Fdownload",{"key":154,"label":155,"href":156},"downloads.piecesOS.macOS.dmgIntel","PiecesOS macOS DMG Intel","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fos_server\u002Fdmg\u002Fdownload",{"key":158,"label":159,"href":160},"downloads.piecesOS.macOS.universal","PiecesOS macOS Universal","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fos_server\u002Fmacos-universal\u002Fdownload",{"key":162,"label":163,"href":164},"downloads.piecesOS.macOS.pkgArm64","PiecesOS macOS PKG Apple Silicon","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fmacos_packaging\u002Fpkg-pos-launch-only-arm64\u002Fdownload",{"key":166,"label":167,"href":168},"downloads.piecesOS.macOS.pkgIntel","PiecesOS macOS PKG Intel","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fmacos_packaging\u002Fpkg-pos-launch-only\u002Fdownload",{"key":170,"label":171,"href":172},"downloads.piecesOS.windows.appinstaller","PiecesOS Windows App Installer","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fappinstaller\u002Fos_server.appinstaller",{"key":174,"label":175,"href":176},"downloads.piecesOS.windows.exe","PiecesOS Windows EXE","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fos_server\u002Fwindows-exe\u002Fdownload",{"key":178,"label":179,"href":180},"downloads.piecesOS.windows.msix","PiecesOS Windows MSIX","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fos_server\u002Fwindows-msix\u002Fdownload",{"key":182,"label":183,"href":184},"downloads.piecesOS.linux.snap","PiecesOS Linux Snap package","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fos_server\u002Fsnap\u002Fdownload",{"key":186,"label":187,"href":188},"downloads.combined.macOS.appleSilicon","Combined macOS installer Apple Silicon","https:\u002F\u002Fbuilds.pieces.app\u002Fstages\u002Fproduction\u002Fmacos_packaging\u002Fpkg-arm64\u002Fdownload",{"key":190,"label":191,"href":104},"downloads.combined.macOS.intel","Combined macOS installer Intel",{"key":193,"label":194,"href":195},"downloads.guides.macOS","macOS installation guide","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces\u002Fmacos-installation-guide",{"key":197,"label":198,"href":199},"downloads.guides.windows","Windows installation guide","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces\u002Fwindows-installation-guide",{"key":201,"label":202,"href":203},"downloads.guides.linux","Linux installation guide","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces\u002Flinux-installation-guide",{"key":205,"label":206,"href":207},"downloads.guides.piecesOS","PiecesOS manual installation","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fcore-dependencies\u002Fpieces-os\u002Fmanual-installation",{"key":209,"label":210,"href":211},"extensions.chrome","Chrome extension","https:\u002F\u002Fchrome.google.com\u002Fwebstore\u002Fdetail\u002Fpieces-save-code-snippets\u002Figbgibhbfonhmjlechmeefimncpekepm",{"key":213,"label":214,"href":215},"extensions.firefox","Firefox add-on","https:\u002F\u002Faddons.mozilla.org\u002Fen-US\u002Ffirefox\u002Faddon\u002Fpieces-save-code-from-the-web\u002F",{"key":217,"label":218,"href":219},"extensions.edge","Edge add-on","https:\u002F\u002Fmicrosoftedge.microsoft.com\u002Faddons\u002Fdetail\u002Fpieces-save-code-snippet\u002Fhglfimcdgonaeeobjckfdabcldfidmim",{"key":221,"label":222,"href":223},"extensions.vscode","VS Code extension","https:\u002F\u002Fmarketplace.visualstudio.com\u002Fitems?itemName=MeshIntelligentTechnologiesInc.pieces-vscode",{"key":225,"label":226,"href":227},"extensions.visualStudio","Visual Studio extension","https:\u002F\u002Fmarketplace.visualstudio.com\u002Fitems?itemName=MeshIntelligentTechnologiesInc.PiecesVisualStudio",{"key":229,"label":230,"href":231},"extensions.jetbrains","JetBrains plugin","https:\u002F\u002Fplugins.jetbrains.com\u002Fplugin\u002F17328-pieces--save-search-share--reuse-code-snippets",{"key":233,"label":234,"href":235},"extensions.obsidian","Obsidian plugin","https:\u002F\u002Fobsidian.md\u002Fplugins?id=pieces-for-developers",{"key":237,"label":238,"href":239},"extensions.sublime","Sublime package","https:\u002F\u002Fpackagecontrol.io\u002Fpackages\u002FPieces",{"key":241,"label":242,"href":243},"extensions.neovim","Neovim plugin","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fplugin_neo_vim",{"key":245,"label":246,"href":247},"extensions.jupyterlab","JupyterLab plugin","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fjupyterlab-pieces",{"key":249,"label":250,"href":251},"extensions.cli","Pieces CLI","https:\u002F\u002Fpypi.org\u002Fproject\u002Fpieces-cli\u002F",{"key":253,"label":254,"href":255},"docs.home","Documentation home","https:\u002F\u002Fdocs.pieces.app",{"key":257,"label":258,"href":259},"docs.getStarted","Get started docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces",{"key":261,"label":262,"href":263},"docs.api","API docs","https:\u002F\u002Fdocs.pieces.app\u002Fapi",{"key":265,"label":266,"href":267},"docs.desktop.overview","Desktop overview","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop",{"key":269,"label":270,"href":271},"docs.desktop.onboarding","Desktop onboarding","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Fonboarding",{"key":273,"label":274,"href":275},"docs.desktop.timeline","Desktop timeline docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Ftimeline",{"key":277,"label":278,"href":279},"docs.desktop.summaries","Desktop summaries docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Fsingle-click-summaries",{"key":281,"label":282,"href":283},"docs.desktop.search","Desktop conversational search docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Fconversational-search",{"key":285,"label":286,"href":287},"docs.desktop.drive","Desktop drive docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Fdrive",{"key":289,"label":290,"href":291},"docs.desktop.account","Desktop account settings docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fdesktop\u002Fconfiguration\u002Faccount",{"key":293,"label":294,"href":92},"docs.desktop.download","Desktop download docs",{"key":296,"label":297,"href":298},"docs.piecesOS.overview","PiecesOS overview docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fcore-dependencies",{"key":300,"label":301,"href":302},"docs.piecesOS.details","PiecesOS details docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fcore-dependencies\u002Fpieces-os",{"key":304,"label":305,"href":207},"docs.piecesOS.install","PiecesOS install docs",{"key":307,"label":308,"href":309},"docs.piecesOS.quickMenu","PiecesOS quick menu docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fcore-dependencies\u002Fpieces-os\u002Fquick-menu",{"key":311,"label":312,"href":313},"docs.piecesOS.storage","On-device storage docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fcore-dependencies\u002Fon-device-storage",{"key":315,"label":316,"href":317},"docs.piecesOS.troubleshooting","PiecesOS troubleshooting docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fcore-dependencies\u002Fpieces-os\u002Ftroubleshooting",{"key":319,"label":320,"href":321},"docs.mcp.overview","MCP overview docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp",{"key":323,"label":324,"href":325},"docs.mcp.cursor","MCP Cursor docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fcursor",{"key":327,"label":328,"href":329},"docs.mcp.vscode","MCP VS Code docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fvs-code",{"key":331,"label":332,"href":333},"docs.mcp.claudeDesktop","MCP Claude Desktop docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fclaude-desktop",{"key":335,"label":336,"href":337},"docs.mcp.claudeCode","MCP Claude Code docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fclaude-code",{"key":339,"label":340,"href":341},"docs.mcp.claudeCowork","MCP Claude Cowork docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fclaude-cowork",{"key":343,"label":344,"href":345},"docs.mcp.githubCopilot","MCP GitHub Copilot docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fgithub-copilot",{"key":347,"label":348,"href":349},"docs.mcp.goose","MCP Goose docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fgoose",{"key":351,"label":352,"href":353},"docs.mcp.windsurf","MCP Windsurf docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fwindsurf",{"key":355,"label":356,"href":357},"docs.mcp.zed","MCP Zed docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fzed",{"key":359,"label":360,"href":361},"docs.mcp.jetbrains","MCP JetBrains docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fjetbrains-ides",{"key":363,"label":364,"href":365},"docs.mcp.continueDev","MCP Continue docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fcontinue-dev",{"key":367,"label":368,"href":369},"docs.mcp.cline","MCP Cline docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fcline",{"key":371,"label":372,"href":373},"docs.mcp.raycast","MCP Raycast docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fraycast",{"key":375,"label":376,"href":377},"docs.mcp.rovoDevCli","MCP Rovo Dev CLI docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Frovo-dev-cli",{"key":379,"label":380,"href":381},"docs.mcp.openaiCodexCli","MCP OpenAI Codex CLI docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fopenai-codex-cli",{"key":383,"label":384,"href":385},"docs.mcp.googleGeminiCli","MCP Google Gemini CLI docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fgoogle-gemini-cli",{"key":387,"label":388,"href":389},"docs.mcp.amazonQ","MCP Amazon Q docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Famazon-q-developer",{"key":391,"label":392,"href":393},"docs.mcp.chatgptDev","MCP ChatGPT Developer Mode docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fchatgpt-developer-mode",{"key":395,"label":396,"href":397},"docs.mcp.openclaw","MCP OpenClaw docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fopenclaw",{"key":399,"label":400,"href":401},"docs.mcp.mcpRemote","MCP Remote docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fmcp-remote",{"key":403,"label":404,"href":405},"docs.mcp.ngrok","MCP ngrok docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmcp\u002Fngrok-setup",{"key":407,"label":408,"href":409},"docs.troubleshooting.macOS","macOS troubleshooting docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces\u002Ftroubleshooting\u002Fmacos",{"key":411,"label":412,"href":413},"docs.troubleshooting.windows","Windows troubleshooting docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces\u002Ftroubleshooting\u002Fwindows",{"key":415,"label":416,"href":417},"docs.troubleshooting.linux","Linux troubleshooting docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fmeet-pieces\u002Ftroubleshooting\u002Flinux",{"key":419,"label":420,"href":421},"docs.privacy","Privacy and security docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fprivacy-security-your-data",{"key":423,"label":424,"href":425},"docs.support","Support docs","https:\u002F\u002Fdocs.pieces.app\u002Fproducts\u002Fsupport",{"key":427,"label":428,"href":429},"portal.home","Pieces portal","https:\u002F\u002Fportal.pieces.app",{"key":431,"label":432,"href":433},"site.home","Website home","https:\u002F\u002Fpieces.app",{"key":435,"label":436,"href":437},"site.about","About page","https:\u002F\u002Fpieces.app\u002Fabout",{"key":439,"label":440,"href":441},"site.features","Features page","https:\u002F\u002Fpieces.app\u002Ffeatures",{"key":443,"label":444,"href":445},"site.plugins","Plugins page","https:\u002F\u002Fpieces.app\u002Fplugins",{"key":447,"label":448,"href":449},"site.contact","Contact page","https:\u002F\u002Fpieces.app\u002Fcontact",{"key":451,"label":58,"href":452},"site.updates","https:\u002F\u002Fpieces.app\u002Fupdates",{"key":454,"label":455,"href":456},"site.news","News","https:\u002F\u002Fpieces.app\u002Fnews",{"key":458,"label":459,"href":460},"site.events","Community events","https:\u002F\u002Fpieces.app\u002Fcommunity\u002Fevents",{"key":462,"label":463,"href":464},"site.userStories","User stories","https:\u002F\u002Fpieces.app\u002Fuser-stories",{"key":466,"label":467,"href":468},"site.academy","Academy","https:\u002F\u002Fpieces.app\u002Flearn\u002Facademy",{"key":470,"label":471,"href":472},"site.support","Website support","https:\u002F\u002Fpieces.app\u002Fsupport",{"key":474,"label":475,"href":476},"site.standup","Standup","https:\u002F\u002Fpieces.app\u002Fstandup",{"key":478,"label":34,"href":479},"site.blog","https:\u002F\u002Fcode.pieces.app\u002Fblog",{"key":481,"label":43,"href":482},"social.discord","https:\u002F\u002Fdiscord.gg\u002Fgetpieces",{"key":484,"label":46,"href":485},"social.x","https:\u002F\u002Fx.com\u002Fgetpieces",{"key":487,"label":488,"href":485},"social.twitter","Twitter",{"key":490,"label":491,"href":492},"social.instagram","Instagram","https:\u002F\u002Fwww.instagram.com\u002Fgetpieces\u002F",{"key":494,"label":495,"href":496},"social.tiktok","TikTok","https:\u002F\u002Fwww.tiktok.com\u002F@getpieces",{"key":498,"label":49,"href":499},"social.linkedin","https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002Fgetpieces\u002F",{"key":501,"label":502,"href":503},"social.youtube","YouTube","https:\u002F\u002Fyoutube.com\u002F@getpieces",{"key":505,"label":506,"href":507},"github.org","GitHub organization","https:\u002F\u002Fgithub.com\u002Fpieces-app",{"key":509,"label":510,"href":511},"github.support","GitHub support","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fsupport",{"key":513,"label":514,"href":515},"github.issues","GitHub issues","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fsupport\u002Fissues",{"key":517,"label":518,"href":519},"github.discussions","GitHub discussions","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fsupport\u002Fdiscussions",{"key":521,"label":522,"href":523},"github.documentation","GitHub documentation","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fdocumentation",{"key":525,"label":526,"href":527},"github.opensource","GitHub open source","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fopensource",{"key":529,"label":530,"href":531},"github.sdks.python","Python SDK","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fpieces-os-client-sdk-for-python",{"key":533,"label":534,"href":535},"github.sdks.typescript","TypeScript SDK","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fpieces-os-client-sdk-for-typescript",{"key":537,"label":538,"href":539},"github.sdks.dart","Dart SDK","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fpieces-os-client-sdk-for-dart",{"key":541,"label":542,"href":543},"github.sdks.kotlin","Kotlin SDK","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fpieces-os-client-sdk-for-kotlin",{"key":545,"label":546,"href":547},"github.plugins.obsidian","Obsidian plugin repository","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fobsidian-pieces",{"key":549,"label":550,"href":247},"github.plugins.jupyterlab","JupyterLab plugin repository",{"key":552,"label":553,"href":554},"github.plugins.sublime","Sublime plugin repository","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fplugin_sublime",{"key":556,"label":557,"href":558},"github.plugins.neovim","Neovim plugin repository","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fplugin_neovim",{"key":560,"label":561,"href":562},"github.cliAgent","CLI agent repository","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fcli-agent",{"key":564,"label":565,"href":566},"github.mcpDart","MCP Dart repository","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fmcp_dart",{"key":568,"label":569,"href":570},"github.awesomePieces","Awesome Pieces repository","https:\u002F\u002Fgithub.com\u002Fpieces-app\u002Fawesome-pieces",{"key":572,"label":573,"href":66},"legal.privacyPolicy","Privacy policy",{"key":575,"label":576,"href":69},"legal.refundPolicy","Refund policy",{"key":578,"label":579,"href":72},"legal.terms","Terms",{"key":581,"label":582,"href":583},"legal.security","Legal security","https:\u002F\u002Fpieces.app\u002Flegal\u002Fsecurity",{"key":585,"label":586,"href":503},"videos.youtubeChannel","YouTube channel",{"key":588,"label":589,"href":590},"videos.gettingStartedDesktop","Getting started desktop video","https:\u002F\u002Fyoutu.be\u002FdUr1lRM_TYk",{"key":592,"label":593,"href":594},"videos.snippetDiscoveryxR","Snippet discovery video","https:\u002F\u002Fyoutu.be\u002FG6vb1USw-30",{"key":596,"label":597,"href":598},"sales.bookACall","Book a sales call","https:\u002F\u002Fcalendar.app.google\u002FWVUDtUfNy5Vst3sH7",{"key":600,"label":601,"href":602},"sales.enterprise","Enterprise form","https:\u002F\u002Fgetpieces.typeform.com\u002Fto\u002FaVQFTvpE",{"key":604,"label":605,"href":519},"sales.feedback","Feedback discussions",{"key":607,"label":608,"href":609},"sales.earlyAccess","Early access form","https:\u002F\u002Fgetpieces.typeform.com\u002Fearlyaccess",{"key":611,"label":612,"href":613},"sales.supportEmail","Support email","mailto:support@pieces.app",{"key":615,"label":616,"href":19},"routes.home","Home route",{"key":618,"label":619,"href":56},"routes.about","About route",{"key":621,"label":622,"href":12},"routes.downloads","Downloads route",{"key":624,"label":625,"href":626},"routes.thanks","Post-install thanks route","\u002Finstall\u002Fthanks",{"key":628,"label":629,"href":630},"routes.pricing","Pricing route","\u002Fpricing",{"key":632,"label":633,"href":634},"cta.freeTrial","Start free trial CTA (Pieces Pro) — used by \u002Fcampaigns\u002F* landing pages","https:\u002F\u002Fcampaigns.pieces.app\u002F",{"key":636,"label":637,"href":22},"routes.enterprise","Enterprise route",{"key":639,"label":640,"href":62},"routes.contact","Contact route",{"key":642,"label":643,"href":59},"routes.updates","Updates route",{"key":645,"label":646,"href":647},"routes.migrationError","Migration error route","\u002Fmigration\u002Ferror",{"key":649,"label":650,"href":651},"routes.authenticated","Post-login authenticated route","\u002Fauth\u002Fsigned-in",{"key":653,"label":654,"href":655},"routes.signed-out","Signed out route","\u002Fauth\u002Fsigned-out",{"key":657,"label":658,"href":659},"routes.authentication-error","Authentication error route","\u002Fauth\u002Ferror",{},"data\u002Fshared\u002Furls","NLkVE1lIUws518K-1QsSB6T-GvlXYIzcRL1HDhqarGQ",{"id":664,"title":665,"body":666,"category":15,"date":1702,"description":1703,"draft":1704,"extension":1705,"icon":1706,"iconColor":1707,"iconTone":1708,"image":1709,"imageAlt":1708,"meta":1710,"navigation":32,"ogImage":1708,"ogImageAlt":1708,"path":1711,"relatedResources":1708,"seo":1712,"stem":1713,"subtitle":1714,"summary":1708,"__hash__":1715},"updates\u002Fupdates\u002Fnano-models.md","Nano-Models power LTM-2.5",{"type":667,"value":668,"toc":1680},"minimark",[669,690,693,696,699,702,705,708,711,716,728,812,816,822,825,830,833,941,945,948,988,991,995,998,1002,1019,1023,1040,1044,1061,1072,1076,1083,1213,1217,1220,1224,1227,1340,1345,1371,1376,1401,1405,1408,1505,1510,1527,1532,1555,1576,1580,1585,1592,1597,1606,1611,1623,1628,1648,1652,1665,1668,1671,1674,1677],[670,671,672,673,680,681,685,686,689],"p",{},"In the pursuit of building ",[674,675,679],"a",{"href":676,"rel":677},"https:\u002F\u002Fpieces.app\u002Ffeatures\u002Flong-term-memory",[678],"nofollow","long-term Artificial Memory"," at the OS level, understanding ",[682,683,684],"em",{},"when"," a user wants to retrieve information is just as crucial as ",[682,687,688],{},"what"," they want.",[670,691,692],{},"In the early days, every step of that retrieval pipeline, from intent classification through span extraction, normalization, enrichment, relevance scoring, formatting, and upload, ran through cloud-hosted LLMs.",[670,694,695],{},"That meant 8–11 preprocessing tasks before touching the memory store, another 2–4 post-processing tasks afterward, and finally a round-trip to a remote model to compose the answer.",[670,697,698],{},"The result?",[670,700,701],{},"Cumulative latency that drags time-to-first-token into the seconds, accuracy hurdles at each stage, user data exposed in transit, and token bills that balloon with every query.",[670,703,704],{},"Our breakthrough with LTM-2.5: two purpose-built on-device nano-models that offload temporal understanding entirely to local hardware — one for interpreting the user's temporal intent, the other for extracting the precise time span(s) implied by their language.",[670,706,707],{},"These specialized models are the result of extensive knowledge distillation from larger foundation models, quantized and pruned to run efficiently on consumer hardware.",[670,709,710],{},"Now, the entire 10–15 step pipeline lives on-device, preserving privacy, slashing costs, and taking the deterministic retrieval of long-term context down from seconds to milliseconds in latency.",[712,713,715],"h2",{"id":714},"when-to-leverage-the-temporal-model","When to leverage the temporal model",[670,717,718,719,723,724,727],{},"Our pipeline depends on two critical steps: determining ",[720,721,722],"strong",{},"intent"," first, then generating one or more ",[720,725,726],{},"time ranges"," representative of the user's natural-language query:",[729,730,731,748],"table",{},[732,733,734],"thead",{},[735,736,737,743],"tr",{},[738,739,740],"th",{},[720,741,742],{},"Use Case",[738,744,745],{},[720,746,747],{},"Description",[749,750,751,762,772,782,792,802],"tbody",{},[735,752,753,759],{},[754,755,756],"td",{},[720,757,758],{},"Content Retrieval",[754,760,761],{},"Fetching past events (\"What was I working on just now?\")",[735,763,764,769],{},[754,765,766],{},[720,767,768],{},"Action \u002F Scheduling",[754,770,771],{},"Setting reminders or appointments (\"Remind me in two hours\")",[735,773,774,779],{},[754,775,776],{},[720,777,778],{},"Future Information \u002F Planning",[754,780,781],{},"Forecasting or \"next week\" inquiries (\"What am I doing tomorrow afternoon?\")",[735,783,784,789],{},[754,785,786],{},[720,787,788],{},"Current Status",[754,790,791],{},"Real-time checks (\"What am I doing right now?\")",[735,793,794,799],{},[754,795,796],{},[720,797,798],{},"Temporal – General",[754,800,801],{},"Ambiguous or loosely specified time references (\"Show me last week around Friday evening\")",[735,803,804,809],{},[754,805,806],{},[720,807,808],{},"Non-Temporal",[754,810,811],{},"Queries without a time component (\"Explain the concept of recursion.\")",[712,813,815],{"id":814},"temporal-range-generation","Temporal range generation",[670,817,818,819,821],{},"Once we've determined that a query requires temporal memory access, we need to precisely identify ",[720,820,684],{}," to search in the user's activity timeline.",[670,823,824],{},"This is where our second nano-model comes into play:",[826,827,829],"h3",{"id":828},"range-types-and-boundaries","Range types and boundaries",[670,831,832],{},"The temporal span predictor handles several distinct types of time references:",[729,834,835,859],{},[732,836,837],{},[735,838,839,844,849,854],{},[738,840,841],{},[720,842,843],{},"Range Type",[738,845,846],{},[720,847,848],{},"Example Query",[738,850,851],{},[720,852,853],{},"Generated Span",[738,855,856],{},[720,857,858],{},"Search Strategy",[749,860,861,877,893,909,925],{},[735,862,863,868,871,874],{},[754,864,865],{},[720,866,867],{},"Point-in-time",[754,869,870],{},"\"Show me what I was doing at 2pm yesterday\"",[754,872,873],{},"Single timestamp with narrow context window",[754,875,876],{},"Precise timestamp lookup with small buffer",[735,878,879,884,887,890],{},[754,880,881],{},[720,882,883],{},"Explicit period",[754,885,886],{},"\"What emails did I receive between Monday and Wednesday?\"",[754,888,889],{},"Clearly defined start\u002Fend boundaries",[754,891,892],{},"Bounded range search with exact limits",[735,894,895,900,903,906],{},[754,896,897],{},[720,898,899],{},"Implicit period",[754,901,902],{},"\"What was I working on last week?\"",[754,904,905],{},"Inferred start\u002Fend based on cultural\u002Fcontextual norms",[754,907,908],{},"Automatically expanded to appropriate calendar boundaries",[735,910,911,916,919,922],{},[754,912,913],{},[720,914,915],{},"Relative recent",[754,917,918],{},"\"What was I just doing?\"",[754,920,921],{},"Short window counting backward from current time",[754,923,924],{},"Recency-prioritized retrieval with adaptive timespan",[735,926,927,932,935,938],{},[754,928,929],{},[720,930,931],{},"Fuzzy historical",[754,933,934],{},"\"Show me that article I read about quantum computing last summer\"",[754,936,937],{},"Broad date range with lower confidence boundaries",[754,939,940],{},"Expanded search space with relevance decay at boundaries",[826,942,944],{"id":943},"optimizing-the-temporal-search-space","Optimizing the temporal search space",[670,946,947],{},"The model doesn't just identify time boundaries — it also generates crucial metadata about search strategy:",[949,950,951,958,973,979],"ul",{},[952,953,954,957],"li",{},[720,955,956],{},"Confidence scores"," for timespan boundaries (enabling better retrieval when dates are ambiguous)",[952,959,960,963,964,968,969,972],{},[720,961,962],{},"Periodicity hints"," for recurring events (distinguishing ",[965,966,967],"code",{},"\"my Monday meeting\""," from ",[965,970,971],{},"\"last Monday's meeting\"",")",[952,974,975,978],{},[720,976,977],{},"Time-zone awareness"," for properly interpreting references when users travel",[952,980,981,984,985,972],{},[720,982,983],{},"Contextual weighting"," that prioritizes activity density over raw timestamps (e.g., for ",[965,986,987],{},"\"when I was working on the Smith project\"",[670,989,990],{},"This specialized temporal range extraction eliminates the need to scan the entire memory corpus for each query, dramatically reducing both computational load and latency while improving retrieval precision.",[712,992,994],{"id":993},"intention-differentiation-edge-cases","Intention differentiation & edge cases",[670,996,997],{},"Ensuring we route queries correctly between retrieval and planning:",[826,999,1001],{"id":1000},"retrieval-vs-planning","Retrieval vs. planning",[949,1003,1004,1012],{},[952,1005,1006,1009,1010],{},[965,1007,1008],{},"\"What was I working on just now?\""," → ",[720,1011,758],{},[952,1013,1014,1009,1017],{},[965,1015,1016],{},"\"What am I doing tomorrow afternoon?\"",[720,1018,778],{},[826,1020,1022],{"id":1021},"broad-vs-specific","Broad vs. specific",[949,1024,1025,1033],{},[952,1026,1027,1009,1030,1032],{},[965,1028,1029],{},"\"Show me last week around Friday evening\"",[720,1031,758],{}," with a loose span",[952,1034,1035,1009,1038],{},[965,1036,1037],{},"\"Plan my weekend for next Friday evening\"",[720,1039,778],{},[826,1041,1043],{"id":1042},"temporal-vs-non-temporal","Temporal vs. non-temporal",[949,1045,1046,1053],{},[952,1047,1048,1009,1051],{},[965,1049,1050],{},"\"What was the website I was just looking at?\"",[720,1052,758],{},[952,1054,1055,1009,1058,1060],{},[965,1056,1057],{},"\"Explain the concept of recursion.\"",[720,1059,808],{}," (no memory lookup)",[670,1062,1063,1064,1067,1068,1071],{},"By clearly distinguishing ",[720,1065,1066],{},"temporal retrieval"," (pulling historical context) from ",[720,1069,1070],{},"temporal reference"," (scheduling or future-oriented intent), our on-device pipeline avoids misrouted cloud calls, cuts latency to the millisecond level, and maintains top-tier accuracy without sacrificing privacy or incurring hidden costs.",[712,1073,1075],{"id":1074},"examples-scenarios","Examples & scenarios",[670,1077,1078,1079,1082],{},"Below are representative user queries, each fed into our pipeline along with the user's local time in UTC (e.g. ",[965,1080,1081],{},"2025-04-17T16:43:02.151857+00:00","):",[1084,1085,1086,1120,1155,1184],"ol",{},[952,1087,1088,1091],{},[720,1089,1090],{},"Recent Activity Retrieval",[949,1092,1093,1099,1105,1114],{},[952,1094,1095,1098],{},[720,1096,1097],{},"Query:"," \"Could you tell me what I was just doing?\"",[952,1100,1101,1104],{},[720,1102,1103],{},"Classifier (23 ms):"," Content Retrieval",[952,1106,1107,1110,1111],{},[720,1108,1109],{},"Span Predictor (102 ms):"," ",[965,1112,1113],{},"2025-04-17T16:37:05.603Z – 2025-04-17T16:43:02.151857Z",[952,1115,1116,1119],{},[720,1117,1118],{},"Showcases:"," precise on-device extraction of the last few minutes of activity",[952,1121,1122,1125],{},[720,1123,1124],{},"Future Planning (Nuanced Task)",[949,1126,1127,1132,1138,1146],{},[952,1128,1129,1131],{},[720,1130,1097],{}," \"I will go to the store tomorrow.\"",[952,1133,1134,1137],{},[720,1135,1136],{},"Classifier (21 ms):"," Temporal – General",[952,1139,1140,1110,1143],{},[720,1141,1142],{},"Span Predictor:",[682,1144,1145],{},"N\u002FA",[952,1147,1148,1150,1151,1154],{},[720,1149,1118],{}," correctly ",[720,1152,1153],{},"not"," generating a past time-range for future intentions—an essential nuance",[952,1156,1157,1160],{},[720,1158,1159],{},"\"Just\" Retrieval Consistency",[949,1161,1162,1167,1172,1179],{},[952,1163,1164,1166],{},[720,1165,1097],{}," \"What was the website I was just looking at?\"",[952,1168,1169,1104],{},[720,1170,1171],{},"Classifier (22 ms):",[952,1173,1174,1110,1177],{},[720,1175,1176],{},"Span Predictor (108 ms):",[965,1178,1113],{},[952,1180,1181,1183],{},[720,1182,1118],{}," consistent span output across semantically similar \"just\" queries",[952,1185,1186,1189],{},[720,1187,1188],{},"Long-Range Historical Query",[949,1190,1191,1196,1200,1208],{},[952,1192,1193,1195],{},[720,1194,1097],{}," \"What was I working on last year around Thanksgiving?\"",[952,1197,1198,1104],{},[720,1199,1171],{},[952,1201,1202,1110,1205],{},[720,1203,1204],{},"Span Predictor (88 ms):",[965,1206,1207],{},"2024-11-01T00:00:00Z – 2024-11-30T23:59:59.999999Z",[952,1209,1210,1212],{},[720,1211,1118],{}," broad date-range generation for loosely specified historical periods",[712,1214,1216],{"id":1215},"benchmarks","Benchmarks",[670,1218,1219],{},"Tested on an Apple M1 Max (32 GB) under heavy load (30+ tabs, video, IDEs, messaging) to simulate real-world conditions:",[826,1221,1223],{"id":1222},"classification-results","Classification results",[670,1225,1226],{},"This table compares how well each model identifies the correct temporal intent label for a given sample.",[729,1228,1229,1263],{},[732,1230,1231],{},[735,1232,1233,1238,1243,1248,1253,1258],{},[738,1234,1235],{},[720,1236,1237],{},"Model Name",[738,1239,1240],{},[720,1241,1242],{},"Accuracy",[738,1244,1245],{},[720,1246,1247],{},"F1 (W)",[738,1249,1250],{},[720,1251,1252],{},"Prec (W)",[738,1254,1255],{},[720,1256,1257],{},"Recall (W)",[738,1259,1260],{},[720,1261,1262],{},"Samples\u002FSec",[749,1264,1265,1283,1302,1321],{},[735,1266,1267,1270,1273,1275,1278,1280],{},[754,1268,1269],{},"nano-temporal-intent (TIME Intent)",[754,1271,1272],{},"0.9930",[754,1274,1272],{},[754,1276,1277],{},"0.9931",[754,1279,1272],{},[754,1281,1282],{},"544.41",[735,1284,1285,1288,1291,1294,1297,1299],{},[754,1286,1287],{},"gemini-1.5-flash-002",[754,1289,1290],{},"0.8241",[754,1292,1293],{},"0.8384",[754,1295,1296],{},"0.8834",[754,1298,1290],{},[754,1300,1301],{},"9.14",[735,1303,1304,1307,1310,1313,1316,1318],{},[754,1305,1306],{},"gpt-4o",[754,1308,1309],{},"0.8634",[754,1311,1312],{},"0.8470",[754,1314,1315],{},"0.8698",[754,1317,1309],{},[754,1319,1320],{},"9.40",[735,1322,1323,1326,1329,1332,1335,1337],{},[754,1324,1325],{},"meta-llama\u002FLlama-3.2-3B-Instruct",[754,1327,1328],{},"0.4604",[754,1330,1331],{},"0.4094",[754,1333,1334],{},"0.4080",[754,1336,1328],{},[754,1338,1339],{},"92.43",[670,1341,1342],{},[720,1343,1344],{},"Legend: Classification Models",[949,1346,1347,1353,1359,1365],{},[952,1348,1349,1352],{},[720,1350,1351],{},"nano-temporal-intent (TIME Intent):"," Our on-device nano-model for intent classification—ultra-lightweight and lightning-fast inference.",[952,1354,1355,1358],{},[720,1356,1357],{},"gemini-1.5-flash-002:"," Google's mid-tier large language model via API; good accuracy but higher latency and cost.",[952,1360,1361,1364],{},[720,1362,1363],{},"gpt-4o:"," OpenAI's flagship multimodal LLM; strong performance at premium compute and pricing.",[952,1366,1367,1370],{},[720,1368,1369],{},"meta-llama\u002FLlama-3.2-3B-Instruct:"," A 3 billion-parameter open-weights LLM; lower accuracy but faster than cloud LLMs.",[670,1372,1373],{},[720,1374,1375],{},"Legend: Classification Metrics",[949,1377,1378,1384,1395],{},[952,1379,1380,1383],{},[720,1381,1382],{},"Accuracy:"," Proportion of samples for which the top-prediction matches the true class.",[952,1385,1386,1388,1389,1388,1391,1394],{},[720,1387,1247],{},", ",[720,1390,1252],{},[720,1392,1393],{},"Recall (W):"," Weighted F1-score, precision, and recall across all intent classes (accounts for class imbalances).",[952,1396,1397,1400],{},[720,1398,1399],{},"Samples\u002FSec:"," Number of inference calls the model can process per second. (higher is better)",[712,1402,1404],{"id":1403},"span-prediction-results","Span prediction results",[670,1406,1407],{},"This table measures how precisely each model extracts the correct time-span from text.",[729,1409,1410,1437],{},[732,1411,1412],{},[735,1413,1414,1418,1423,1428,1433],{},[738,1415,1416],{},[720,1417,1237],{},[738,1419,1420],{},[720,1421,1422],{},"E.C.O. Rate",[738,1424,1425],{},[720,1426,1427],{},"Avg IoU",[738,1429,1430],{},[720,1431,1432],{},"Exact Match",[738,1434,1435],{},[720,1436,1262],{},[749,1438,1439,1456,1473,1489],{},[735,1440,1441,1444,1447,1450,1453],{},[754,1442,1443],{},"nano-temporal-span-pred (TIME Range)",[754,1445,1446],{},"0.9450",[754,1448,1449],{},"0.9201",[754,1451,1452],{},"0.8659",[754,1454,1455],{},"785.39",[735,1457,1458,1461,1464,1467,1470],{},[754,1459,1460],{},"gemini-1.5-pro-002",[754,1462,1463],{},"0.2065",[754,1465,1466],{},"0.1865",[754,1468,1469],{},"0.1684",[754,1471,1472],{},"9.35",[735,1474,1475,1477,1480,1483,1486],{},[754,1476,1306],{},[754,1478,1479],{},"0.1767",[754,1481,1482],{},"0.1611",[754,1484,1485],{},"0.1535",[754,1487,1488],{},"9.47",[735,1490,1491,1493,1496,1499,1502],{},[754,1492,1325],{},[754,1494,1495],{},"0.1725",[754,1497,1498],{},"0.1640",[754,1500,1501],{},"0.1517",[754,1503,1504],{},"62.02",[670,1506,1507],{},[720,1508,1509],{},"Legend: Span Models",[949,1511,1512,1518],{},[952,1513,1514,1517],{},[720,1515,1516],{},"nano-temporal-span-pred (TIME Range):"," On-device span extractor optimized for low latency and high IoU.",[952,1519,1520,1388,1522,1388,1524,1526],{},[720,1521,1460],{},[720,1523,1306],{},[720,1525,1369],{}," LLMs & SLMs performing span extraction via API calls.",[670,1528,1529],{},[720,1530,1531],{},"Legend: Span Metrics",[949,1533,1534,1539,1544,1550],{},[952,1535,1536,1538],{},[720,1537,1422],{}," (Exact Coverage Overlap): Fraction of predicted spans that exactly match the gold span boundaries.",[952,1540,1541,1543],{},[720,1542,1427],{}," (Intersection-over-Union): Average overlap ratio between predicted and true spans.",[952,1545,1546,1549],{},[720,1547,1548],{},"Exact Match:"," Strict percentage of samples where predicted span text equals ground truth.",[952,1551,1552,1554],{},[720,1553,1399],{}," Span-prediction throughput on the benchmark hardware. (higher is better)",[1556,1557,1558],"blockquote",{},[670,1559,1560,1563,1564,1567,1568,1571,1572,1575],{},[720,1561,1562],{},"We observed SLMs running in the cloud"," on H100 GPU with vLLM incur $0.018 – $1.90 per run and took 15-25 min of compute time — our cascade delivers structured time-spans offline in ",[720,1565,1566],{},"milliseconds",", with ",[720,1569,1570],{},"zero API cost"," and ",[720,1573,1574],{},"full data privacy",".",[712,1577,1579],{"id":1578},"why-it-matters","Why it matters",[670,1581,1582],{},[720,1583,1584],{},"🏗️ Architectural Specialization",[670,1586,1587,1588,1591],{},"Breaking monolithic LLMs into nano-models for classification vs. span prediction yields ",[720,1589,1590],{},"massive gains"," in both accuracy and speed.",[670,1593,1594],{},[720,1595,1596],{},"🌐 Edge-First AI",[670,1598,1599,1600,1605],{},"Offline inference ",[674,1601,1604],{"href":1602,"rel":1603},"https:\u002F\u002Fpieces.app\u002Fblog\u002Foffline-ai",[678],"keeps sensitive data on-device"," — critical for medical, defense, and privacy-focused applications.",[670,1607,1608],{},[720,1609,1610],{},"💡 Energy & Cost Efficiency",[670,1612,1613,1618,1619,1622],{},[674,1614,1617],{"href":1615,"rel":1616},"https:\u002F\u002Fpieces.app\u002Fblog\u002Fsmall-language-models-outshine-large-language-models-enterprise-users",[678],"Eliminate token fees"," and slash compute budgets. This is the future of ",[720,1620,1621],{},"sustainable",", scaled AI on laptops, wearables, and IoT.",[670,1624,1625],{},[720,1626,1627],{},"🔬 Research Frontiers",[949,1629,1630,1636,1642],{},[952,1631,1632,1635],{},[720,1633,1634],{},"Task-specific:"," distillation, quantization, and final pruning for modular pipelines",[952,1637,1638,1641],{},[720,1639,1640],{},"Adaptive orchestration",": dynamic model selection based on compute availability",[952,1643,1644,1647],{},[720,1645,1646],{},"Hardware\u002Fsoftware:"," co-design for ultra-efficient inference",[712,1649,1651],{"id":1650},"conclusion","Conclusion",[670,1653,1654,1655,1388,1658,1661,1662,1575],{},"This nano-temporal pipeline is one of approximately 11 nano-models we're weaving into LTM-2.5 to make long-term memory formation and retrieval across your entire OS ",[720,1656,1657],{},"blazingly fast",[720,1659,1660],{},"highly accurate",", and ",[720,1663,1664],{},"privacy-first",[670,1666,1667],{},"Innovation isn't about bigger models — it's about smarter, specialized models that deliver tangible benefits in real-world applications.",[670,1669,1670],{},"By focusing on modular, purpose-built AI systems that run entirely on-device, we're redefining what's possible for intelligent, responsive computing that respects user privacy while dramatically reducing cost and latency.",[670,1672,1673],{},"We can't wait to share more as we push the boundaries of on-device AI in the world of OS-level Long-Term Memory.",[670,1675,1676],{},"Lastly, I would be remiss if I didn't mention the obvious: none of this would be possible without the incredible creativity, dedication, and perseverance from the team behind Pieces.",[670,1678,1679],{},"I’ll close with a special shout out to our ML team and a extra special shout out to Antreas Antoniou and Sam Jones for believing in the approach and turning these first-principal theories into breakthroughs ✨",{"title":1681,"searchDepth":1682,"depth":1682,"links":1683},"",2,[1684,1685,1690,1695,1696,1699,1700,1701],{"id":714,"depth":1682,"text":715},{"id":814,"depth":1682,"text":815,"children":1686},[1687,1689],{"id":828,"depth":1688,"text":829},3,{"id":943,"depth":1688,"text":944},{"id":993,"depth":1682,"text":994,"children":1691},[1692,1693,1694],{"id":1000,"depth":1688,"text":1001},{"id":1021,"depth":1688,"text":1022},{"id":1042,"depth":1688,"text":1043},{"id":1074,"depth":1682,"text":1075},{"id":1215,"depth":1682,"text":1216,"children":1697},[1698],{"id":1222,"depth":1688,"text":1223},{"id":1403,"depth":1682,"text":1404},{"id":1578,"depth":1682,"text":1579},{"id":1650,"depth":1682,"text":1651},"2025-04-17T00:00:00.000Z","In the pursuit of building long-term Artificial Memory at the OS level, understanding when a user wants to retrieve information is just as crucial as what they want.",false,"md","material-symbols:memory-rounded","#D500F9",null,"https:\u002F\u002Fstorage.googleapis.com\u002Fpieces-marketing-website\u002Fimages\u002Fblog\u002Fnano-models\u002Fhero.png",{},"\u002Fupdates\u002Fnano-models",{"title":665,"description":1703},"updates\u002Fnano-models","Discover the latest breakthrough in AI with Nano-Models as the Pieces Team unveils LTM‑2.5. A game-changer for AI-powered development!","_teYoSnPg1HhTqrrunf02Au1y8nXEmksX3CxD3xBR0U",[],{"left":1718,"top":1718,"width":1719,"height":1719,"rotate":1718,"vFlip":1704,"hFlip":1704,"body":1720},0,24,"\u003Cpath fill=\"currentColor\" d=\"m7.825 13l4.9 4.9q.3.3.288.7t-.313.7q-.3.275-.7.288t-.7-.288l-6.6-6.6q-.15-.15-.213-.325T4.426 12t.063-.375t.212-.325l6.6-6.6q.275-.275.688-.275t.712.275q.3.3.3.713t-.3.712L7.825 11H19q.425 0 .713.288T20 12t-.288.713T19 13z\"\u002F>"]