[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-qdrant-qdrant-model-migration":3,"mdc-4h8ixr-key":48,"related-repo-qdrant-qdrant-model-migration":526,"related-org-qdrant-qdrant-model-migration":626},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":23,"repoUrl":24,"updatedAt":25,"license":26,"forks":27,"topics":28,"repo":43,"sourceUrl":46,"mdContent":47},"qdrant-model-migration","migrate embedding models in Qdrant","Guides embedding model migration in Qdrant without downtime. Use when someone asks 'how to switch embedding models', 'how to migrate vectors', 'how to update to a new model', 'zero-downtime model change', 'how to re-embed my data', or 'can I use two models at once'. Also use when upgrading model dimensions, switching providers, or A\u002FB testing models.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"qdrant","Qdrant","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fqdrant.png",[12,16,17,20],{"name":13,"slug":14,"type":15},"Performance","performance","tag",{"name":9,"slug":8,"type":15},{"name":18,"slug":19,"type":15},"Database","database",{"name":21,"slug":22,"type":15},"Migration","migration",197,"https:\u002F\u002Fgithub.com\u002Fqdrant\u002Fskills","2026-07-16T06:00:48.841055",null,23,[29,30,31,32,33,34,35,36,37,14,8,38,39,40,41,42],"agent-skills","ai-agents","claude-code","codex","cursor","embeddings","hybrid-search","monitoring","multitenancy","quantization","scaling","search-quality","vector-database","vector-search",{"repoUrl":24,"stars":23,"forks":27,"topics":44,"description":45},[29,30,31,32,33,34,35,36,37,14,8,38,39,40,41,42],"Agent skills for Qdrant vector search: scaling, performance optimization, search quality, monitoring, deployment, model migration, version upgrades, and SDK usage across Python, TypeScript, Rust, Go, .NET, Java","https:\u002F\u002Fgithub.com\u002Fqdrant\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fqdrant-model-migration","---\nname: qdrant-model-migration\ndescription: \"Guides embedding model migration in Qdrant without downtime. Use when someone asks 'how to switch embedding models', 'how to migrate vectors', 'how to update to a new model', 'zero-downtime model change', 'how to re-embed my data', or 'can I use two models at once'. Also use when upgrading model dimensions, switching providers, or A\u002FB testing models.\"\n---\n\n# What to Do When Changing Embedding Models\n\nVectors from different models are incompatible. You cannot mix old and new embeddings in the same vector space. On v1.18+, you can add or delete named vector fields on an existing collection — migration no longer always requires a new collection. On v1.17 or earlier, all named vectors must be defined at collection creation time.\n\n- Understand collection aliases before choosing a strategy [Collection aliases](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=collection-aliases)\n\n\n## Can I Avoid Re-embedding?\n\nUse when: looking for shortcuts before committing to full migration.\n\nYou MUST re-embed if: changing model provider (OpenAI to Cohere), changing architecture (CLIP to BGE), incompatible dimension counts across different models, or adding sparse vectors to dense-only collection.\n\nYou CAN avoid re-embedding if: using Matryoshka models (use `dimensions` parameter to output lower-dimensional embeddings, learn linear transformation from sample data, some recall loss, good for 100M+ datasets). Or changing quantization (binary to scalar): Qdrant re-quantizes automatically. [Quantization](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fquantization\u002F)\n\n\n## Need Zero Downtime\n\nUse when: production must stay available. Recommended for model replacement at scale.\n\n- If the cluster is v1.18 or later AND the collection has named vectors:\n\n  - Add the new vector field directly to the existing collection [Update vector schema](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=update-vector-schema)\n  - Re-embed all data in the background using `UpdateVectors` [Update vectors](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fpoints\u002F?s=update-vectors)\n  - Verify search quality, then delete old vector field\n\n- If the cluster is v1.17 or earlier OR the collection doesn't have named vectors:\n\n- Create a new collection with the new model's dimensions and distance metric\n- Re-embed all data into the new collection in the background\n- Point your application at a collection alias instead of a direct collection name\n- Atomically swap the alias to the new collection [Switch collection](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=switch-collection)\n- Verify search quality, then delete the old collection\n\nCareful, the alias swap only redirects queries. Payloads must be re-uploaded separately.\n\n\n## Need Both Models Live (Side-by-Side)\n\nUse when: A\u002FB testing models, multi-modal (dense + sparse), or evaluating a new model before committing.\n\n- If the cluster is v1.18 or later:\n\n  - Add the new vector field directly to the existing collection [Update vector schema](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=update-vector-schema)\n  - Backfill new model embeddings incrementally using `UpdateVectors` [Update vectors](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fpoints\u002F?s=update-vectors)\n\n- If the cluster is v1.17 or earlier: You cannot add a named vector to an existing collection. Create a new collection with both vector fields defined upfront:\n\n  - Create new collection with old and new named vectors both defined [Collection with multiple vectors](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=collection-with-multiple-vectors)\n  - Migrate data from old collection, preserving existing vectors in the old named field\n  - Backfill new model embeddings incrementally using `UpdateVectors` [Update vectors](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fpoints\u002F?s=update-vectors)\n  - Compare quality by querying with `using: \"old_model\"` vs `using: \"new_model\"`\n  - Swap alias to new collection once satisfied\n\nCo-locating large multi-vectors (especially ColBERT) with dense vectors degrades ALL queries, even those only using dense. At millions of points, users report 13s latency dropping to 2s after removing ColBERT. Put large vectors on disk during side-by-side migration.\n\nIf you anticipate future model migrations, define both vector fields upfront at collection creation.\n\n\n## Dense to Hybrid Search Migration\n\nUse when: adding sparse\u002FBM25 vectors to an existing dense-only collection. Most common migration pattern.\n\nYou cannot add sparse vectors to an existing collection that uses a default (unnamed) dense vector. Must recreate:\n\n- Create new collection with both dense and sparse vector configs defined\n- Re-embed all data with both dense and sparse models\n- Migrate payloads, swap alias\n\nIf the collection already uses named dense vectors and is on v1.18+, add the sparse vector field directly without recreating [Update vector schema](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=update-vector-schema).\n\nSparse vectors at chunk level have different TF-IDF characteristics than document level. Test retrieval quality after migration, especially for non-English text without stop-word removal.\n\n\n## Re-embedding Is Too Slow\n\nUse when: dataset is large and re-embedding is the bottleneck.\n\n- Use `update_mode: insert` (v1.17+) for safe idempotent migration [Update mode](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fpoints\u002F?s=update-mode)\n- Scroll the old collection with `with_vectors=False`, re-embed in batches, upsert into new collection\n- Upload in parallel batches (64-256 points per request, 2-4 parallel streams) [Bulk upload](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fbulk-upload\u002F)\n- Disable HNSW during bulk load (set `indexing_threshold_kb` very high, restore after)\n- For Qdrant Cloud inference, switching models is a config change, not a pipeline change [Inference docs](https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Finference\u002F)\n\nFor 400GB+ datasets, expect days. For small datasets (\u003C25MB), re-indexing from source is faster than using the migration tool.\n\n\n## What NOT to Do\n\n- Assume you can add named vectors to an existing collection on v1.17 or earlier servers; check your server version first\n- Delete the old collection before verifying the new one\n- Forget to update the query embedding model in your application code\n- Skip payload migration when using alias swap (aliases redirect queries, they do not copy data)\n- Keep ColBERT vectors co-located with dense vectors during a long migration (I\u002FO cost degrades all queries)\n- Migrate to hybrid search without testing BM25 quality at chunk level\n",{"data":49,"body":50},{"name":4,"description":6},{"type":51,"children":52},"root",[53,62,68,87,94,99,104,125,131,136,221,226,232,237,337,342,347,353,358,363,381,393,398,404,409,482,487,493],{"type":54,"tag":55,"props":56,"children":58},"element","h1",{"id":57},"what-to-do-when-changing-embedding-models",[59],{"type":60,"value":61},"text","What to Do When Changing Embedding Models",{"type":54,"tag":63,"props":64,"children":65},"p",{},[66],{"type":60,"value":67},"Vectors from different models are incompatible. You cannot mix old and new embeddings in the same vector space. On v1.18+, you can add or delete named vector fields on an existing collection — migration no longer always requires a new collection. On v1.17 or earlier, all named vectors must be defined at collection creation time.",{"type":54,"tag":69,"props":70,"children":71},"ul",{},[72],{"type":54,"tag":73,"props":74,"children":75},"li",{},[76,78],{"type":60,"value":77},"Understand collection aliases before choosing a strategy ",{"type":54,"tag":79,"props":80,"children":84},"a",{"href":81,"rel":82},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=collection-aliases",[83],"nofollow",[85],{"type":60,"value":86},"Collection aliases",{"type":54,"tag":88,"props":89,"children":91},"h2",{"id":90},"can-i-avoid-re-embedding",[92],{"type":60,"value":93},"Can I Avoid Re-embedding?",{"type":54,"tag":63,"props":95,"children":96},{},[97],{"type":60,"value":98},"Use when: looking for shortcuts before committing to full migration.",{"type":54,"tag":63,"props":100,"children":101},{},[102],{"type":60,"value":103},"You MUST re-embed if: changing model provider (OpenAI to Cohere), changing architecture (CLIP to BGE), incompatible dimension counts across different models, or adding sparse vectors to dense-only collection.",{"type":54,"tag":63,"props":105,"children":106},{},[107,109,116,118],{"type":60,"value":108},"You CAN avoid re-embedding if: using Matryoshka models (use ",{"type":54,"tag":110,"props":111,"children":113},"code",{"className":112},[],[114],{"type":60,"value":115},"dimensions",{"type":60,"value":117}," parameter to output lower-dimensional embeddings, learn linear transformation from sample data, some recall loss, good for 100M+ datasets). Or changing quantization (binary to scalar): Qdrant re-quantizes automatically. ",{"type":54,"tag":79,"props":119,"children":122},{"href":120,"rel":121},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fquantization\u002F",[83],[123],{"type":60,"value":124},"Quantization",{"type":54,"tag":88,"props":126,"children":128},{"id":127},"need-zero-downtime",[129],{"type":60,"value":130},"Need Zero Downtime",{"type":54,"tag":63,"props":132,"children":133},{},[134],{"type":60,"value":135},"Use when: production must stay available. Recommended for model replacement at scale.",{"type":54,"tag":69,"props":137,"children":138},{},[139,184,189,194,199,204,216],{"type":54,"tag":73,"props":140,"children":141},{},[142,144],{"type":60,"value":143},"If the cluster is v1.18 or later AND the collection has named vectors:",{"type":54,"tag":69,"props":145,"children":146},{},[147,159,179],{"type":54,"tag":73,"props":148,"children":149},{},[150,152],{"type":60,"value":151},"Add the new vector field directly to the existing collection ",{"type":54,"tag":79,"props":153,"children":156},{"href":154,"rel":155},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=update-vector-schema",[83],[157],{"type":60,"value":158},"Update vector schema",{"type":54,"tag":73,"props":160,"children":161},{},[162,164,170,172],{"type":60,"value":163},"Re-embed all data in the background using ",{"type":54,"tag":110,"props":165,"children":167},{"className":166},[],[168],{"type":60,"value":169},"UpdateVectors",{"type":60,"value":171}," ",{"type":54,"tag":79,"props":173,"children":176},{"href":174,"rel":175},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fpoints\u002F?s=update-vectors",[83],[177],{"type":60,"value":178},"Update vectors",{"type":54,"tag":73,"props":180,"children":181},{},[182],{"type":60,"value":183},"Verify search quality, then delete old vector field",{"type":54,"tag":73,"props":185,"children":186},{},[187],{"type":60,"value":188},"If the cluster is v1.17 or earlier OR the collection doesn't have named vectors:",{"type":54,"tag":73,"props":190,"children":191},{},[192],{"type":60,"value":193},"Create a new collection with the new model's dimensions and distance metric",{"type":54,"tag":73,"props":195,"children":196},{},[197],{"type":60,"value":198},"Re-embed all data into the new collection in the background",{"type":54,"tag":73,"props":200,"children":201},{},[202],{"type":60,"value":203},"Point your application at a collection alias instead of a direct collection name",{"type":54,"tag":73,"props":205,"children":206},{},[207,209],{"type":60,"value":208},"Atomically swap the alias to the new collection ",{"type":54,"tag":79,"props":210,"children":213},{"href":211,"rel":212},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=switch-collection",[83],[214],{"type":60,"value":215},"Switch collection",{"type":54,"tag":73,"props":217,"children":218},{},[219],{"type":60,"value":220},"Verify search quality, then delete the old collection",{"type":54,"tag":63,"props":222,"children":223},{},[224],{"type":60,"value":225},"Careful, the alias swap only redirects queries. Payloads must be re-uploaded separately.",{"type":54,"tag":88,"props":227,"children":229},{"id":228},"need-both-models-live-side-by-side",[230],{"type":60,"value":231},"Need Both Models Live (Side-by-Side)",{"type":54,"tag":63,"props":233,"children":234},{},[235],{"type":60,"value":236},"Use when: A\u002FB testing models, multi-modal (dense + sparse), or evaluating a new model before committing.",{"type":54,"tag":69,"props":238,"children":239},{},[240,273],{"type":54,"tag":73,"props":241,"children":242},{},[243,245],{"type":60,"value":244},"If the cluster is v1.18 or later:",{"type":54,"tag":69,"props":246,"children":247},{},[248,257],{"type":54,"tag":73,"props":249,"children":250},{},[251,252],{"type":60,"value":151},{"type":54,"tag":79,"props":253,"children":255},{"href":154,"rel":254},[83],[256],{"type":60,"value":158},{"type":54,"tag":73,"props":258,"children":259},{},[260,262,267,268],{"type":60,"value":261},"Backfill new model embeddings incrementally using ",{"type":54,"tag":110,"props":263,"children":265},{"className":264},[],[266],{"type":60,"value":169},{"type":60,"value":171},{"type":54,"tag":79,"props":269,"children":271},{"href":174,"rel":270},[83],[272],{"type":60,"value":178},{"type":54,"tag":73,"props":274,"children":275},{},[276,278],{"type":60,"value":277},"If the cluster is v1.17 or earlier: You cannot add a named vector to an existing collection. Create a new collection with both vector fields defined upfront:",{"type":54,"tag":69,"props":279,"children":280},{},[281,293,298,313,332],{"type":54,"tag":73,"props":282,"children":283},{},[284,286],{"type":60,"value":285},"Create new collection with old and new named vectors both defined ",{"type":54,"tag":79,"props":287,"children":290},{"href":288,"rel":289},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fcollections\u002F?s=collection-with-multiple-vectors",[83],[291],{"type":60,"value":292},"Collection with multiple vectors",{"type":54,"tag":73,"props":294,"children":295},{},[296],{"type":60,"value":297},"Migrate data from old collection, preserving existing vectors in the old named field",{"type":54,"tag":73,"props":299,"children":300},{},[301,302,307,308],{"type":60,"value":261},{"type":54,"tag":110,"props":303,"children":305},{"className":304},[],[306],{"type":60,"value":169},{"type":60,"value":171},{"type":54,"tag":79,"props":309,"children":311},{"href":174,"rel":310},[83],[312],{"type":60,"value":178},{"type":54,"tag":73,"props":314,"children":315},{},[316,318,324,326],{"type":60,"value":317},"Compare quality by querying with ",{"type":54,"tag":110,"props":319,"children":321},{"className":320},[],[322],{"type":60,"value":323},"using: \"old_model\"",{"type":60,"value":325}," vs ",{"type":54,"tag":110,"props":327,"children":329},{"className":328},[],[330],{"type":60,"value":331},"using: \"new_model\"",{"type":54,"tag":73,"props":333,"children":334},{},[335],{"type":60,"value":336},"Swap alias to new collection once satisfied",{"type":54,"tag":63,"props":338,"children":339},{},[340],{"type":60,"value":341},"Co-locating large multi-vectors (especially ColBERT) with dense vectors degrades ALL queries, even those only using dense. At millions of points, users report 13s latency dropping to 2s after removing ColBERT. Put large vectors on disk during side-by-side migration.",{"type":54,"tag":63,"props":343,"children":344},{},[345],{"type":60,"value":346},"If you anticipate future model migrations, define both vector fields upfront at collection creation.",{"type":54,"tag":88,"props":348,"children":350},{"id":349},"dense-to-hybrid-search-migration",[351],{"type":60,"value":352},"Dense to Hybrid Search Migration",{"type":54,"tag":63,"props":354,"children":355},{},[356],{"type":60,"value":357},"Use when: adding sparse\u002FBM25 vectors to an existing dense-only collection. Most common migration pattern.",{"type":54,"tag":63,"props":359,"children":360},{},[361],{"type":60,"value":362},"You cannot add sparse vectors to an existing collection that uses a default (unnamed) dense vector. Must recreate:",{"type":54,"tag":69,"props":364,"children":365},{},[366,371,376],{"type":54,"tag":73,"props":367,"children":368},{},[369],{"type":60,"value":370},"Create new collection with both dense and sparse vector configs defined",{"type":54,"tag":73,"props":372,"children":373},{},[374],{"type":60,"value":375},"Re-embed all data with both dense and sparse models",{"type":54,"tag":73,"props":377,"children":378},{},[379],{"type":60,"value":380},"Migrate payloads, swap alias",{"type":54,"tag":63,"props":382,"children":383},{},[384,386,391],{"type":60,"value":385},"If the collection already uses named dense vectors and is on v1.18+, add the sparse vector field directly without recreating ",{"type":54,"tag":79,"props":387,"children":389},{"href":154,"rel":388},[83],[390],{"type":60,"value":158},{"type":60,"value":392},".",{"type":54,"tag":63,"props":394,"children":395},{},[396],{"type":60,"value":397},"Sparse vectors at chunk level have different TF-IDF characteristics than document level. Test retrieval quality after migration, especially for non-English text without stop-word removal.",{"type":54,"tag":88,"props":399,"children":401},{"id":400},"re-embedding-is-too-slow",[402],{"type":60,"value":403},"Re-embedding Is Too Slow",{"type":54,"tag":63,"props":405,"children":406},{},[407],{"type":60,"value":408},"Use when: dataset is large and re-embedding is the bottleneck.",{"type":54,"tag":69,"props":410,"children":411},{},[412,432,445,457,470],{"type":54,"tag":73,"props":413,"children":414},{},[415,417,423,425],{"type":60,"value":416},"Use ",{"type":54,"tag":110,"props":418,"children":420},{"className":419},[],[421],{"type":60,"value":422},"update_mode: insert",{"type":60,"value":424}," (v1.17+) for safe idempotent migration ",{"type":54,"tag":79,"props":426,"children":429},{"href":427,"rel":428},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fpoints\u002F?s=update-mode",[83],[430],{"type":60,"value":431},"Update mode",{"type":54,"tag":73,"props":433,"children":434},{},[435,437,443],{"type":60,"value":436},"Scroll the old collection with ",{"type":54,"tag":110,"props":438,"children":440},{"className":439},[],[441],{"type":60,"value":442},"with_vectors=False",{"type":60,"value":444},", re-embed in batches, upsert into new collection",{"type":54,"tag":73,"props":446,"children":447},{},[448,450],{"type":60,"value":449},"Upload in parallel batches (64-256 points per request, 2-4 parallel streams) ",{"type":54,"tag":79,"props":451,"children":454},{"href":452,"rel":453},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Fmanage-data\u002Fbulk-upload\u002F",[83],[455],{"type":60,"value":456},"Bulk upload",{"type":54,"tag":73,"props":458,"children":459},{},[460,462,468],{"type":60,"value":461},"Disable HNSW during bulk load (set ",{"type":54,"tag":110,"props":463,"children":465},{"className":464},[],[466],{"type":60,"value":467},"indexing_threshold_kb",{"type":60,"value":469}," very high, restore after)",{"type":54,"tag":73,"props":471,"children":472},{},[473,475],{"type":60,"value":474},"For Qdrant Cloud inference, switching models is a config change, not a pipeline change ",{"type":54,"tag":79,"props":476,"children":479},{"href":477,"rel":478},"https:\u002F\u002Fskills.qdrant.tech\u002Fmd\u002Fdocumentation\u002Finference\u002F",[83],[480],{"type":60,"value":481},"Inference docs",{"type":54,"tag":63,"props":483,"children":484},{},[485],{"type":60,"value":486},"For 400GB+ datasets, expect days. For small datasets (\u003C25MB), re-indexing from source is faster than using the migration tool.",{"type":54,"tag":88,"props":488,"children":490},{"id":489},"what-not-to-do",[491],{"type":60,"value":492},"What NOT to Do",{"type":54,"tag":69,"props":494,"children":495},{},[496,501,506,511,516,521],{"type":54,"tag":73,"props":497,"children":498},{},[499],{"type":60,"value":500},"Assume you can add named vectors to an existing collection on v1.17 or earlier servers; check your server version first",{"type":54,"tag":73,"props":502,"children":503},{},[504],{"type":60,"value":505},"Delete the old collection before verifying the new one",{"type":54,"tag":73,"props":507,"children":508},{},[509],{"type":60,"value":510},"Forget to update the query embedding model in your application code",{"type":54,"tag":73,"props":512,"children":513},{},[514],{"type":60,"value":515},"Skip payload migration when using alias swap (aliases redirect queries, they do not copy data)",{"type":54,"tag":73,"props":517,"children":518},{},[519],{"type":60,"value":520},"Keep ColBERT vectors co-located with dense vectors during a long migration (I\u002FO cost degrades all queries)",{"type":54,"tag":73,"props":522,"children":523},{},[524],{"type":60,"value":525},"Migrate to hybrid search without testing BM25 quality at chunk level",{"items":527,"total":625},[528,543,558,573,588,602,615],{"slug":529,"name":529,"fn":530,"description":531,"org":532,"tags":533,"stars":23,"repoUrl":24,"updatedAt":542},"qdrant-advisor","diagnose and troubleshoot Qdrant deployments","Diagnose, troubleshoot, and advise on any Qdrant deployment by loading the latest official Qdrant skills live from skills.qdrant.tech. Use this whenever someone raises a Qdrant problem or question — slow or degraded search, high or growing memory \u002F OOM crashes, optimizer stuck or slow, indexing slowness, scaling and sharding decisions (node count, QPS, latency, multitenancy, vertical vs horizontal), poor or irrelevant search results, hybrid search and reranking, embedding-model migration, version upgrades and compatibility, monitoring and observability (Prometheus, Grafana, health checks, \u002Fmetrics, \u002Ftelemetry), deployment choices (local, Docker, self-hosted, Qdrant Cloud, embedded), or client-SDK questions (Python, TypeScript, Rust, Go, .NET, Java). Trigger especially when the context is clearly a Qdrant cluster, collection, or vector-search deployment. Always prefer this skill over answering from memory: it pulls current, authoritative guidance and only the relevant context.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[534,535,538,541],{"name":18,"slug":19,"type":15},{"name":536,"slug":537,"type":15},"Debugging","debugging",{"name":539,"slug":540,"type":15},"Operations","operations",{"name":9,"slug":8,"type":15},"2026-07-16T06:02:56.675293",{"slug":544,"name":544,"fn":545,"description":546,"org":547,"tags":548,"stars":23,"repoUrl":24,"updatedAt":557},"qdrant-clients-sdk","integrate Qdrant client SDKs","Qdrant provides client SDKs for various programming languages, allowing easy integration with Qdrant deployments.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[549,552,553,554],{"name":550,"slug":551,"type":15},"API Development","api-development",{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":555,"slug":556,"type":15},"SDK","sdk","2026-07-16T05:59:57.799789",{"slug":559,"name":559,"fn":560,"description":561,"org":562,"tags":563,"stars":23,"repoUrl":24,"updatedAt":572},"qdrant-deployment-options","select Qdrant deployment options","Guides Qdrant deployment selection. Use when someone asks 'how to deploy Qdrant', 'Docker vs Cloud', 'local mode', 'embedded Qdrant', 'Qdrant EDGE', 'which deployment option', 'self-hosted vs cloud', or 'need lowest latency deployment'. Also use when choosing between deployment types for a new project.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[564,565,568,571],{"name":18,"slug":19,"type":15},{"name":566,"slug":567,"type":15},"Deployment","deployment",{"name":569,"slug":570,"type":15},"Infrastructure","infrastructure",{"name":9,"slug":8,"type":15},"2026-07-19T05:38:33.324265",{"slug":574,"name":574,"fn":575,"description":576,"org":577,"tags":578,"stars":23,"repoUrl":24,"updatedAt":587},"qdrant-edge","build applications with Qdrant Edge","Guides building on Qdrant Edge, the embedded in-process shard. Use when someone asks 'how to sync Edge with the server', 'keep a local shard in sync with Qdrant Cloud', 'BM25 or keyword search on Edge', 'hybrid search on Edge', 'embeddings on device', 'Edge snapshots', 'apply a partial snapshot', 'why is my Edge search empty after inserts', or is writing custom sync, BM25, or fusion code against qdrant-edge. Also use when deciding what Edge ships built-in versus what you must implement.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[579,580,583,584],{"name":18,"slug":19,"type":15},{"name":581,"slug":582,"type":15},"Edge","edge",{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},"Search","search","2026-07-16T06:01:19.179105",{"slug":589,"name":589,"fn":590,"description":591,"org":592,"tags":593,"stars":23,"repoUrl":24,"updatedAt":601},"qdrant-horizontal-scaling","guide Qdrant horizontal scaling decisions","Diagnoses and guides Qdrant horizontal scaling decisions. Use when someone asks 'vertical or horizontal?', 'how many nodes?', 'how many shards?', 'how to add nodes', 'resharding', 'data doesn't fit', or 'need more capacity'. Also use when data growth outpaces current deployment.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[594,597,598,599],{"name":595,"slug":596,"type":15},"Architecture","architecture",{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":600,"slug":39,"type":15},"Scaling","2026-07-19T05:38:32.348551",{"slug":603,"name":603,"fn":604,"description":605,"org":606,"tags":607,"stars":23,"repoUrl":24,"updatedAt":614},"qdrant-hybrid-search","implement hybrid search in Qdrant","Explains hybrid search in Qdrant. Use when someone asks 'how do I setup hybrid search?', 'how to combine keyword and semantic search?', 'sparse plus dense vectors?', 'missing keyword matches', 'how to combine results from multiple searches?' and 'combining multiple representations'",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[608,611,612,613],{"name":609,"slug":610,"type":15},"AI Infrastructure","ai-infrastructure",{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},"2026-07-16T06:00:12.328122",{"slug":616,"name":616,"fn":617,"description":618,"org":619,"tags":620,"stars":23,"repoUrl":24,"updatedAt":624},"qdrant-hybrid-search-combining","combine hybrid search scores in Qdrant","Fusing scores from multiple searches into a single ranked result (RRF, DBSF, custom fusion). Use when someone asks 'RRF or DBSF?', 'how to combine sparse and dense', 'how to combine scores from multiple searches?', 'custom fusion', or 'fusion is not producing good results'",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[621,622,623],{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},"2026-07-16T06:01:49.401142",30,{"items":627,"total":625},[628,635,642,649,656,663,670,676,687,698,708,719],{"slug":529,"name":529,"fn":530,"description":531,"org":629,"tags":630,"stars":23,"repoUrl":24,"updatedAt":542},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[631,632,633,634],{"name":18,"slug":19,"type":15},{"name":536,"slug":537,"type":15},{"name":539,"slug":540,"type":15},{"name":9,"slug":8,"type":15},{"slug":544,"name":544,"fn":545,"description":546,"org":636,"tags":637,"stars":23,"repoUrl":24,"updatedAt":557},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[638,639,640,641],{"name":550,"slug":551,"type":15},{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":555,"slug":556,"type":15},{"slug":559,"name":559,"fn":560,"description":561,"org":643,"tags":644,"stars":23,"repoUrl":24,"updatedAt":572},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[645,646,647,648],{"name":18,"slug":19,"type":15},{"name":566,"slug":567,"type":15},{"name":569,"slug":570,"type":15},{"name":9,"slug":8,"type":15},{"slug":574,"name":574,"fn":575,"description":576,"org":650,"tags":651,"stars":23,"repoUrl":24,"updatedAt":587},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[652,653,654,655],{"name":18,"slug":19,"type":15},{"name":581,"slug":582,"type":15},{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},{"slug":589,"name":589,"fn":590,"description":591,"org":657,"tags":658,"stars":23,"repoUrl":24,"updatedAt":601},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[659,660,661,662],{"name":595,"slug":596,"type":15},{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":600,"slug":39,"type":15},{"slug":603,"name":603,"fn":604,"description":605,"org":664,"tags":665,"stars":23,"repoUrl":24,"updatedAt":614},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[666,667,668,669],{"name":609,"slug":610,"type":15},{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},{"slug":616,"name":616,"fn":617,"description":618,"org":671,"tags":672,"stars":23,"repoUrl":24,"updatedAt":624},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[673,674,675],{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},{"slug":677,"name":677,"fn":678,"description":679,"org":680,"tags":681,"stars":23,"repoUrl":24,"updatedAt":686},"qdrant-hybrid-search-prefetches","configure hybrid search with prefetch queries","Constructing prefetch queries for hybrid retrieval, including sparse\u002Fdense and multi-field setups, and choosing a sparse embedding model. Use when someone asks 'dense and sparse in one search?', 'how to combine multiple fields for retrieval?', 'payloads or sparse vectors for lexical?', 'which sparse embedding model to use?', or 'BM25 vs SPLADE?'",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[682,683,684,685],{"name":550,"slug":551,"type":15},{"name":18,"slug":19,"type":15},{"name":9,"slug":8,"type":15},{"name":585,"slug":586,"type":15},"2026-07-16T06:03:02.305907",{"slug":688,"name":688,"fn":689,"description":690,"org":691,"tags":692,"stars":23,"repoUrl":24,"updatedAt":697},"qdrant-indexing-performance-optimization","optimize Qdrant indexing and data ingestion","Diagnoses and fixes slow Qdrant indexing and data ingestion. Use when someone reports 'uploads are slow', 'indexing takes forever', 'optimizer is stuck', 'HNSW build time too long', or 'data uploaded but search is bad'. Also use when optimizer status shows errors, segments won't merge, or indexing threshold questions arise.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[693,694,695,696],{"name":18,"slug":19,"type":15},{"name":536,"slug":537,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},"2026-07-19T05:38:36.321999",{"slug":699,"name":699,"fn":700,"description":701,"org":702,"tags":703,"stars":23,"repoUrl":24,"updatedAt":707},"qdrant-memory-usage-optimization","optimize Qdrant memory usage","Diagnoses and reduces Qdrant memory usage. Use when someone reports 'memory too high', 'RAM keeps growing', 'node crashed', 'out of memory', 'memory leak', or asks 'why is memory usage so high?', 'how to reduce RAM?'. Also use when memory doesn't match calculations, quantization didn't help, or nodes crash during recovery.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[704,705,706],{"name":536,"slug":537,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},"2026-07-16T06:00:15.357243",{"slug":709,"name":709,"fn":710,"description":711,"org":712,"tags":713,"stars":23,"repoUrl":24,"updatedAt":718},"qdrant-minimize-latency","optimize Qdrant query latency","Guides Qdrant query latency optimization. Use when someone asks 'search is slow', 'how to reduce latency', 'p99 is too high', 'tail latency', 'single query too slow', 'how to make search faster', or 'latency spikes'.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[714,715,716,717],{"name":18,"slug":19,"type":15},{"name":536,"slug":537,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},"2026-07-16T05:59:17.325839",{"slug":4,"name":4,"fn":5,"description":6,"org":720,"tags":721,"stars":23,"repoUrl":24,"updatedAt":25},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[722,723,724,725],{"name":18,"slug":19,"type":15},{"name":21,"slug":22,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15}]