[{"data":1,"prerenderedAt":93},["ShallowReactive",2],{"$f3qbbjkuyh13v0":3},{"site":4,"app":16,"alternatives":69,"compose":88,"log":89,"changelog":90},{"hourly_rate_usd":5,"break_even_window_mo":6,"reference_env":7,"attribution":8},20,18,"Fresh Debian 12, Docker + Compose preinstalled, 2 GB RAM \u002F 2 vCPU ($5-class VPS)",{"caniselfhostit":9},{"name":10,"url":11,"repo":12,"license":13,"author":14,"note":15},"caniselfhostit.com","https:\u002F\u002Fcaniselfhostit.com","https:\u002F\u002Fgithub.com\u002Fcaniselfhostit\u002Fcaniselfhostit","MIT","Jashanpreet Singh","Capability data (app → replacement pairings, pricing snapshots) seeded from this dataset. They answer 'can you run it, and how' — go there for guided installs.",{"id":17,"name":18,"domain":19,"category":20,"price_usd_mo":21,"price_plan":22,"price_unit":23,"price_source":24,"price_checked":25,"jobs":26,"verdict":32,"verdict_reason":33,"what_you_lose":34,"alternatives":38,"economics":40,"hardware_tier":43,"provenance":44,"verified":47,"compose":55,"links":59,"notes":61,"derived":62},"chatgpt","ChatGPT","chatgpt.com","ai-assistants",8,"Go","flat","https:\u002F\u002Fopenai.com\u002Fchatgpt\u002Fpricing\u002F","2026-08-05",[27,28,29,30,31],"frontier-model answers","image generation","voice conversations","memory & custom instructions","chat UI with history","NOT_REALLY","LibreChat self-hosts the chat window, not the intelligence. You'll bring API keys (still paying, now per-token) or a GPU whose electricity bill reads like a subscription. The model was the product all along.",[35,36,37],"The frontier model itself — local models on consumer hardware are years behind what the subscription buys","Voice, image generation, and memory unless you wire up (and pay for) each provider separately","Predictable pricing: API metering can beat $8\u002Fmo for light use, or wildly exceed it",[39],"librechat",{"vps_share_usd_mo":41,"storage_usd_mo":42,"maint_min_mo":5},1,0,"old-laptop",{"capability_seed":45,"decision_layer":46},"caniselfhostit","original",{"by":48,"at":49,"protocol":50,"setup_min":51,"assistant":52,"evidence":53,"what_broke":56},"zernonia","2026-08-10","v1",50,"claude-code",[54,55],"logs\u002Fchatgpt-librechat.md","compose\u002Fchatgpt-librechat.yml",[57,58],"Images are ghcr.io-hosted — Docker-Hub-only mirrors can't pull them; this pair's compose boot is verified by the CI runner","Boots healthy with zero API keys configured — which demonstrates the verdict: a working chat UI with nothing to talk to",{"caniselfhostit":60},"https:\u002F\u002Fcaniselfhostit.com\u002Fchatgpt","The KINDA hiding inside: heavy API users with existing keys get a better UI and per-token pricing from LibreChat. For everyone else the subscription is the cheap option.",{"hourly_rate_usd":5,"self_host_cost_usd_mo":63,"net_saving_usd_mo":64,"setup_cost_usd":65,"break_even_months":51,"break_even_date":66,"markup_index":67,"primary_abandoned":68,"verdict_stale":68},7.67,0.33,16.67,"2030-10-10",1.04,false,[70],{"id":39,"name":71,"tagline":72,"repo":73,"site":74,"license":13,"category":75,"resources":76,"provenance":81,"metrics":82},"LibreChat","One chat window in front of Anthropic, OpenAI and Google, with the conversation history in a database you own.","danny-avila\u002FLibreChat","https:\u002F\u002Fwww.librechat.ai","ai-tools",{"ram_min_mb":77,"disk_gb":78,"arm64":79,"containers":80,"external_db":79},2048,10,true,3,{"capability_seed":45,"decision_layer":46},{"id":39,"stars":83,"last_commit":84,"latest_release":85,"docker_pulls":85,"open_issues":86,"fetched_at":87},41899,"2026-08-11T03:54:24Z",null,689,"2026-08-11T03:55:25.354Z","# LibreChat — self-hosted chat UI (tested against ChatGPT's job list; see the verdict for\n# why the UI was never the thing you were paying for)\n# Note: without provider API keys this boots and serves, but answers require either paid API\n# keys (you're still paying, per-token) or a local model server + GPU. That IS the verdict.\nservices:\n  librechat:\n    image: ghcr.io\u002Fdanny-avila\u002Flibrechat:latest\n    container_name: librechat\n    restart: unless-stopped\n    depends_on:\n      mongodb:\n        condition: service_healthy\n    environment:\n      HOST: 0.0.0.0\n      PORT: \"3080\"\n      MONGO_URI: mongodb:\u002F\u002Fmongodb:27017\u002FLibreChat\n      CREDS_KEY: f34be427ebb29de8d88c107a71546019685ed8b241d8f2ed00c3df97ad2566f0\n      CREDS_IV: e2341419ec3dd3d19b13a1a87fafcbfb\n      JWT_SECRET: 16f8c0ef4a5d391b26034086c628469d3f9f497f08163ab9b40137092f2909ef\n      JWT_REFRESH_SECRET: eaa5191f2914e30b9387fd84e254e4ba6fc51b4654968a9b0803b456a54b8418\n    ports:\n      - \"3080:3080\"\n    healthcheck:\n      test: [\"CMD-SHELL\", \"node -e \\\"fetch('http:\u002F\u002Flocalhost:3080\u002Fhealth').then(r=>process.exit(r.ok?0:1)).catch(()=>process.exit(1))\\\"\"]\n      interval: 5s\n      timeout: 5s\n      retries: 36\n      start_period: 30s\n\n  mongodb:\n    image: mongo:7\n    restart: unless-stopped\n    command: mongod --noauth\n    volumes:\n      - librechat-mongo:\u002Fdata\u002Fdb\n    healthcheck:\n      test: [\"CMD-SHELL\", \"mongosh --quiet --eval 'db.runCommand({ping:1}).ok' | grep 1\"]\n      interval: 5s\n      timeout: 5s\n      retries: 24\n      start_period: 15s\n\nvolumes:\n  librechat-mongo:\n","# Timed setup log: ChatGPT → LibreChat\n\n**Protocol:** v1 · **Verified by:** zernonia · **Date:** 2026-08-10\n**Assistant:** claude-code · **Environment:** containerized runner, 2 vCPU class, Docker 29.3 \u002F Compose v5.1\n\n## Timeline\n\n| Step | Time |\n|---|---|\n| Read config docs, draft compose (app + mongo, JWT\u002FCREDS secrets) | 20 min |\n| Boot attempt in our verification environment: **blocked — ghcr.io unreachable** (see What broke) | 5 min |\n| Compose boot + health verified by the CI runner instead (`compose-check` workflow) | — |\n| Workflow review against docs: registration, provider key config, endpoint selection | 15 min |\n| **Total: ~50 min** | |\n\n## Measurements\n\n- Verification environment could not pull ghcr.io images; **boot-to-healthy for this pair is\n  certified by CI** (same `scripts\u002Fcheck-compose.sh` gate, per-PR and weekly).\n- 2 containers (app + mongo); ~700 MB RAM together before any local model enters the picture.\n\n## What broke\n\n1. **ghcr.io-only images** — same note as Plausible CE: Docker-Hub-only mirrors can't pull.\n2. **It boots healthy with zero API keys configured** — which is the verdict in one sentence:\n   you get a working chat UI with nothing to talk to. Every answer still requires either a\n   provider API key (you're paying per-token again) or a local model server plus hardware\n   that can feed it.\n\n## Verdict-relevant notes\n\n- For heavy API users with existing keys, LibreChat is genuinely good — multi-provider, clean\n  UI, your history in your Mongo. That's the KINDA hiding inside the NOT REALLY.\n- For replacing a normal ChatGPT subscription: the model is the product; the UI was never\n  what the money bought.\n",[91],{"date":49,"app":17,"from":85,"to":32,"reason":92},"Initial verdict: you'd self-host the wrapper; the model is the product.",1786459421183]