[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-replicate-compare-models":3,"mdc-qbykgj-key":33,"related-org-replicate-compare-models":306,"related-repo-replicate-compare-models":387},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":22,"repoUrl":23,"updatedAt":24,"license":25,"forks":26,"topics":27,"repo":28,"sourceUrl":31,"mdContent":32},"compare-models","compare Replicate AI models","Compare Replicate models by cost, speed, quality, and capabilities.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"replicate","Replicate","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Freplicate.jpg",[12,16,19],{"name":13,"slug":14,"type":15},"Benchmarking","benchmarking","tag",{"name":17,"slug":18,"type":15},"LLM","llm",{"name":20,"slug":21,"type":15},"Analysis","analysis",51,"https:\u002F\u002Fgithub.com\u002Freplicate\u002Fskills","2026-07-16T06:03:10.401096",null,6,[],{"repoUrl":23,"stars":22,"forks":26,"topics":29,"description":30},[],"A collection of Agent Skills for building AI-powered apps with Replicate","https:\u002F\u002Fgithub.com\u002Freplicate\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fcompare-models","---\nname: compare-models\ndescription: Compare Replicate models by cost, speed, quality, and capabilities.\n---\n\n## Docs\n\n- Reference: \u003Chttps:\u002F\u002Freplicate.com\u002Fdocs\u002Fllms.txt>\n- OpenAPI schema: \u003Chttps:\u002F\u002Fapi.replicate.com\u002Fopenapi.json>\n- MCP server: \u003Chttps:\u002F\u002Fmcp.replicate.com>\n- Per-model docs: `https:\u002F\u002Freplicate.com\u002F{owner}\u002F{model}\u002Fllms.txt`\n- Set `Accept: text\u002Fmarkdown` when requesting docs pages for Markdown responses.\n\n## Workflow\n\n1. Search or browse collections to build a shortlist of candidate models.\n2. Fetch each model's schema to compare inputs, outputs, and capabilities.\n3. Check pricing from model metadata or the Replicate website.\n4. Run a small batch of test predictions to compare output quality.\n5. Pick the model that best fits your constraints (cost, latency, quality).\n\n## What to compare\n\n- **Speed**: Check `metrics.predict_time` on completed predictions for actual inference time. Official models are always warm. Community models can cold-boot.\n- **Cost**: Official models have predictable per-run pricing. Community models charge by compute time (GPU-seconds). Run a few predictions and check the `metrics` field for actual cost.\n- **Quality**: Run the same prompts through each model and compare outputs. Quality is subjective. Match it to your use case, not a leaderboard.\n- **Capabilities**: Compare input schemas for supported features (reference images, masks, aspect ratios, streaming, multi-image input). Check output formats.\n\n## Key tradeoffs\n\n- Lowest cost: smaller\u002Fdistilled models. Accept slower inference and lower quality.\n- Lowest latency: official models or schnell\u002Fturbo variants. Accept higher cost per run.\n- Highest quality: pro\u002Fmax\u002Fquality variants. Accept slower inference and higher cost.\n- Most control: models with ControlNet, masks, or reference images. Accept more complex input setup.\n\n## Official vs community models\n\n- Official models: always warm, stable APIs, predictable pricing, maintained by Replicate.\n- Community models: may cold-boot, require version pinning, maintained by the author.\n- If a community model meets your needs and an official model doesn't, consider creating a deployment for consistent uptime.\n\n## Prompting guidance\n\nFor prompting techniques and task-specific guidance:\n\n- Image generation and editing: see the [prompt-images](..\u002Fprompt-images\u002FSKILL.md) skill.\n- Video generation: see the [prompt-videos](..\u002Fprompt-videos\u002FSKILL.md) skill.\n",{"data":34,"body":35},{"name":4,"description":6},{"type":36,"children":37},"root",[38,47,112,118,147,153,213,219,242,248,266,272,278],{"type":39,"tag":40,"props":41,"children":43},"element","h2",{"id":42},"docs",[44],{"type":45,"value":46},"text","Docs",{"type":39,"tag":48,"props":49,"children":50},"ul",{},[51,65,76,87,99],{"type":39,"tag":52,"props":53,"children":54},"li",{},[55,57],{"type":45,"value":56},"Reference: ",{"type":39,"tag":58,"props":59,"children":63},"a",{"href":60,"rel":61},"https:\u002F\u002Freplicate.com\u002Fdocs\u002Fllms.txt",[62],"nofollow",[64],{"type":45,"value":60},{"type":39,"tag":52,"props":66,"children":67},{},[68,70],{"type":45,"value":69},"OpenAPI schema: ",{"type":39,"tag":58,"props":71,"children":74},{"href":72,"rel":73},"https:\u002F\u002Fapi.replicate.com\u002Fopenapi.json",[62],[75],{"type":45,"value":72},{"type":39,"tag":52,"props":77,"children":78},{},[79,81],{"type":45,"value":80},"MCP server: ",{"type":39,"tag":58,"props":82,"children":85},{"href":83,"rel":84},"https:\u002F\u002Fmcp.replicate.com",[62],[86],{"type":45,"value":83},{"type":39,"tag":52,"props":88,"children":89},{},[90,92],{"type":45,"value":91},"Per-model docs: ",{"type":39,"tag":93,"props":94,"children":96},"code",{"className":95},[],[97],{"type":45,"value":98},"https:\u002F\u002Freplicate.com\u002F{owner}\u002F{model}\u002Fllms.txt",{"type":39,"tag":52,"props":100,"children":101},{},[102,104,110],{"type":45,"value":103},"Set ",{"type":39,"tag":93,"props":105,"children":107},{"className":106},[],[108],{"type":45,"value":109},"Accept: text\u002Fmarkdown",{"type":45,"value":111}," when requesting docs pages for Markdown responses.",{"type":39,"tag":40,"props":113,"children":115},{"id":114},"workflow",[116],{"type":45,"value":117},"Workflow",{"type":39,"tag":119,"props":120,"children":121},"ol",{},[122,127,132,137,142],{"type":39,"tag":52,"props":123,"children":124},{},[125],{"type":45,"value":126},"Search or browse collections to build a shortlist of candidate models.",{"type":39,"tag":52,"props":128,"children":129},{},[130],{"type":45,"value":131},"Fetch each model's schema to compare inputs, outputs, and capabilities.",{"type":39,"tag":52,"props":133,"children":134},{},[135],{"type":45,"value":136},"Check pricing from model metadata or the Replicate website.",{"type":39,"tag":52,"props":138,"children":139},{},[140],{"type":45,"value":141},"Run a small batch of test predictions to compare output quality.",{"type":39,"tag":52,"props":143,"children":144},{},[145],{"type":45,"value":146},"Pick the model that best fits your constraints (cost, latency, quality).",{"type":39,"tag":40,"props":148,"children":150},{"id":149},"what-to-compare",[151],{"type":45,"value":152},"What to compare",{"type":39,"tag":48,"props":154,"children":155},{},[156,175,193,203],{"type":39,"tag":52,"props":157,"children":158},{},[159,165,167,173],{"type":39,"tag":160,"props":161,"children":162},"strong",{},[163],{"type":45,"value":164},"Speed",{"type":45,"value":166},": Check ",{"type":39,"tag":93,"props":168,"children":170},{"className":169},[],[171],{"type":45,"value":172},"metrics.predict_time",{"type":45,"value":174}," on completed predictions for actual inference time. Official models are always warm. Community models can cold-boot.",{"type":39,"tag":52,"props":176,"children":177},{},[178,183,185,191],{"type":39,"tag":160,"props":179,"children":180},{},[181],{"type":45,"value":182},"Cost",{"type":45,"value":184},": Official models have predictable per-run pricing. Community models charge by compute time (GPU-seconds). Run a few predictions and check the ",{"type":39,"tag":93,"props":186,"children":188},{"className":187},[],[189],{"type":45,"value":190},"metrics",{"type":45,"value":192}," field for actual cost.",{"type":39,"tag":52,"props":194,"children":195},{},[196,201],{"type":39,"tag":160,"props":197,"children":198},{},[199],{"type":45,"value":200},"Quality",{"type":45,"value":202},": Run the same prompts through each model and compare outputs. Quality is subjective. Match it to your use case, not a leaderboard.",{"type":39,"tag":52,"props":204,"children":205},{},[206,211],{"type":39,"tag":160,"props":207,"children":208},{},[209],{"type":45,"value":210},"Capabilities",{"type":45,"value":212},": Compare input schemas for supported features (reference images, masks, aspect ratios, streaming, multi-image input). Check output formats.",{"type":39,"tag":40,"props":214,"children":216},{"id":215},"key-tradeoffs",[217],{"type":45,"value":218},"Key tradeoffs",{"type":39,"tag":48,"props":220,"children":221},{},[222,227,232,237],{"type":39,"tag":52,"props":223,"children":224},{},[225],{"type":45,"value":226},"Lowest cost: smaller\u002Fdistilled models. Accept slower inference and lower quality.",{"type":39,"tag":52,"props":228,"children":229},{},[230],{"type":45,"value":231},"Lowest latency: official models or schnell\u002Fturbo variants. Accept higher cost per run.",{"type":39,"tag":52,"props":233,"children":234},{},[235],{"type":45,"value":236},"Highest quality: pro\u002Fmax\u002Fquality variants. Accept slower inference and higher cost.",{"type":39,"tag":52,"props":238,"children":239},{},[240],{"type":45,"value":241},"Most control: models with ControlNet, masks, or reference images. Accept more complex input setup.",{"type":39,"tag":40,"props":243,"children":245},{"id":244},"official-vs-community-models",[246],{"type":45,"value":247},"Official vs community models",{"type":39,"tag":48,"props":249,"children":250},{},[251,256,261],{"type":39,"tag":52,"props":252,"children":253},{},[254],{"type":45,"value":255},"Official models: always warm, stable APIs, predictable pricing, maintained by Replicate.",{"type":39,"tag":52,"props":257,"children":258},{},[259],{"type":45,"value":260},"Community models: may cold-boot, require version pinning, maintained by the author.",{"type":39,"tag":52,"props":262,"children":263},{},[264],{"type":45,"value":265},"If a community model meets your needs and an official model doesn't, consider creating a deployment for consistent uptime.",{"type":39,"tag":40,"props":267,"children":269},{"id":268},"prompting-guidance",[270],{"type":45,"value":271},"Prompting guidance",{"type":39,"tag":273,"props":274,"children":275},"p",{},[276],{"type":45,"value":277},"For prompting techniques and task-specific guidance:",{"type":39,"tag":48,"props":279,"children":280},{},[281,294],{"type":39,"tag":52,"props":282,"children":283},{},[284,286,292],{"type":45,"value":285},"Image generation and editing: see the ",{"type":39,"tag":58,"props":287,"children":289},{"href":288},"..\u002Fprompt-images\u002FSKILL.md",[290],{"type":45,"value":291},"prompt-images",{"type":45,"value":293}," skill.",{"type":39,"tag":52,"props":295,"children":296},{},[297,299,305],{"type":45,"value":298},"Video generation: see the ",{"type":39,"tag":58,"props":300,"children":302},{"href":301},"..\u002Fprompt-videos\u002FSKILL.md",[303],{"type":45,"value":304},"prompt-videos",{"type":45,"value":293},{"items":307,"total":386},[308,324,330,340,353,364,376],{"slug":309,"name":309,"fn":310,"description":311,"org":312,"tags":313,"stars":22,"repoUrl":23,"updatedAt":323},"build-models","build and package custom AI models","Package and build custom AI models with Cog for deployment on Replicate. Use when creating a cog.yaml or predict.py, defining model inputs and outputs, loading model weights at setup time, building Docker images for ML models, serving locally with cog serve or cog predict, or porting a HuggingFace, GitHub, or ComfyUI model to run on Replicate. Trigger on phrases like \"build a model\", \"package a model\", \"create a Cog model\", \"wrap a model\", \"containerize an AI model\", \"predict.py\", \"cog.yaml\", \"BasePredictor\", or \"Cog container\", and when referencing cog.run, github.com\u002Freplicate\u002Fcog, or github.com\u002Freplicate\u002Fcog-examples. Covers GPU and CUDA setup, pget for fast weight downloads, async predictors with continuous batching, streaming outputs, and cold-boot optimization for image, video, audio, and LLM models. For pushing built models to Replicate, see publish-models. For running existing models, see run-models.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[314,317,320],{"name":315,"slug":316,"type":15},"AI Infrastructure","ai-infrastructure",{"name":318,"slug":319,"type":15},"Deployment","deployment",{"name":321,"slug":322,"type":15},"Machine Learning","machine-learning","2026-07-16T06:02:28.318441",{"slug":4,"name":4,"fn":5,"description":6,"org":325,"tags":326,"stars":22,"repoUrl":23,"updatedAt":24},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[327,328,329],{"name":20,"slug":21,"type":15},{"name":13,"slug":14,"type":15},{"name":17,"slug":18,"type":15},{"slug":331,"name":331,"fn":332,"description":333,"org":334,"tags":335,"stars":22,"repoUrl":23,"updatedAt":339},"find-models","find AI models on Replicate","Find AI models on Replicate using search and curated collections.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[336,337,338],{"name":315,"slug":316,"type":15},{"name":17,"slug":18,"type":15},{"name":321,"slug":322,"type":15},"2026-07-16T06:01:52.508562",{"slug":291,"name":291,"fn":341,"description":342,"org":343,"tags":344,"stars":22,"repoUrl":23,"updatedAt":352},"generate prompts for image models","Prompting techniques for AI image generation and editing models on Replicate. Use when writing prompts for image models or building image generation features.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[345,348,351],{"name":346,"slug":347,"type":15},"Creative","creative",{"name":349,"slug":350,"type":15},"Image Generation","image-generation",{"name":17,"slug":18,"type":15},"2026-07-16T06:01:52.867443",{"slug":304,"name":304,"fn":354,"description":355,"org":356,"tags":357,"stars":22,"repoUrl":23,"updatedAt":363},"generate prompts for AI video models","Prompting techniques for AI video generation models on Replicate. Use when writing prompts for video models or building video generation features.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[358,359,360],{"name":346,"slug":347,"type":15},{"name":17,"slug":18,"type":15},{"name":361,"slug":362,"type":15},"Video","video","2026-07-16T06:03:10.065511",{"slug":365,"name":365,"fn":366,"description":367,"org":368,"tags":369,"stars":22,"repoUrl":23,"updatedAt":375},"publish-models","publish and deploy AI models to Replicate","Push and publish custom AI models to Replicate, and set up CI\u002FCD for releasing new model versions safely. Use when running cog push, deploying a model to Replicate, releasing a new version, validating a model with cog-safe-push before publishing, configuring a Replicate deployment, setting up GitHub Actions for model releases, or porting a community model to an official one. Trigger on phrases like \"push a model to Replicate\", \"publish a model\", \"deploy a model\", \"release a new version\", \"cog push\", \"cog-safe-push\", \"model CI\", \"r8.im\", or \"schema compatibility\", and when referencing github.com\u002Freplicate\u002Fcog-safe-push or github.com\u002Freplicate\u002Fmodel-ci-template. Covers cog push, the full cog-safe-push config (test cases, fuzz, deployment, official_model), GitHub Actions patterns, multi-model matrix pushes, and post-publish monitoring. Assumes you already have a working Cog project; see build-models if you need to package one first.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[370,371,374],{"name":315,"slug":316,"type":15},{"name":372,"slug":373,"type":15},"CI\u002FCD","ci-cd",{"name":318,"slug":319,"type":15},"2026-07-16T06:02:28.648245",{"slug":377,"name":377,"fn":378,"description":379,"org":380,"tags":381,"stars":22,"repoUrl":23,"updatedAt":385},"run-models","run AI models on Replicate","Run AI models on Replicate via predictions, webhooks, and streaming.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[382,383,384],{"name":315,"slug":316,"type":15},{"name":17,"slug":18,"type":15},{"name":321,"slug":322,"type":15},"2026-07-16T06:03:09.738392",7,{"items":388,"total":386},[389,395,401,407,413,419,425],{"slug":309,"name":309,"fn":310,"description":311,"org":390,"tags":391,"stars":22,"repoUrl":23,"updatedAt":323},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[392,393,394],{"name":315,"slug":316,"type":15},{"name":318,"slug":319,"type":15},{"name":321,"slug":322,"type":15},{"slug":4,"name":4,"fn":5,"description":6,"org":396,"tags":397,"stars":22,"repoUrl":23,"updatedAt":24},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[398,399,400],{"name":20,"slug":21,"type":15},{"name":13,"slug":14,"type":15},{"name":17,"slug":18,"type":15},{"slug":331,"name":331,"fn":332,"description":333,"org":402,"tags":403,"stars":22,"repoUrl":23,"updatedAt":339},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[404,405,406],{"name":315,"slug":316,"type":15},{"name":17,"slug":18,"type":15},{"name":321,"slug":322,"type":15},{"slug":291,"name":291,"fn":341,"description":342,"org":408,"tags":409,"stars":22,"repoUrl":23,"updatedAt":352},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[410,411,412],{"name":346,"slug":347,"type":15},{"name":349,"slug":350,"type":15},{"name":17,"slug":18,"type":15},{"slug":304,"name":304,"fn":354,"description":355,"org":414,"tags":415,"stars":22,"repoUrl":23,"updatedAt":363},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[416,417,418],{"name":346,"slug":347,"type":15},{"name":17,"slug":18,"type":15},{"name":361,"slug":362,"type":15},{"slug":365,"name":365,"fn":366,"description":367,"org":420,"tags":421,"stars":22,"repoUrl":23,"updatedAt":375},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[422,423,424],{"name":315,"slug":316,"type":15},{"name":372,"slug":373,"type":15},{"name":318,"slug":319,"type":15},{"slug":377,"name":377,"fn":378,"description":379,"org":426,"tags":427,"stars":22,"repoUrl":23,"updatedAt":385},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[428,429,430],{"name":315,"slug":316,"type":15},{"name":17,"slug":18,"type":15},{"name":321,"slug":322,"type":15}]