> Explore Qdrant's agent skills catalog at https://skills.qdrant.tech/ > Search the documentation at https://skills.qdrant.tech/search?query=your+query+here > Use this file to discover all available pages: https://qdrant.tech/llms.txt # FastEmbed Choose a model for the input and vector type you need. [FastEmbed](https://qdrant.tech/documentation/fastembed/index.md) runs embedding models locally in your Python application, so you can send the resulting vectors to any Qdrant deployment. These tables list the built-in embedding models in [FastEmbed 0.9.0](https://github.com/qdrant/fastembed/releases/tag/v0.9.0). Model names, dimensions, and licenses come from that release's model registry. Your installed version determines which models you can use. ## Dense Text Models These models produce one dense vector per text input. Use the listed dimensions when you create a collection. | Model | Dimensions | License | |---|---|---| | `BAAI/bge-base-en` | 768 | mit | | `BAAI/bge-base-en-v1.5` | 768 | mit | | `BAAI/bge-large-en-v1.5` | 1024 | mit | | `BAAI/bge-small-en` | 384 | mit | | `BAAI/bge-small-en-v1.5` | 384 | mit | | `BAAI/bge-small-zh-v1.5` | 512 | mit | | `google/embeddinggemma-300m` | 768 | gemma | | `google/siglip2-base-patch16-224` | 768 | apache-2.0 | | `ibm-granite/granite-embedding-small-english-r2` | 384 | apache-2.0 | | `intfloat/multilingual-e5-large` | 1024 | mit | | `jinaai/jina-clip-v1` | 768 | apache-2.0 | | `jinaai/jina-embeddings-v2-base-code` | 768 | apache-2.0 | | `jinaai/jina-embeddings-v2-base-de` | 768 | apache-2.0 | | `jinaai/jina-embeddings-v2-base-en` | 768 | apache-2.0 | | `jinaai/jina-embeddings-v2-base-es` | 768 | apache-2.0 | | `jinaai/jina-embeddings-v2-base-zh` | 768 | apache-2.0 | | `jinaai/jina-embeddings-v2-small-en` | 512 | apache-2.0 | | `jinaai/jina-embeddings-v3` | 1024 | cc-by-nc-4.0 | | `minishlab/potion-base-8M` | 256 | mit | | `minishlab/potion-multilingual-128M` | 256 | mit | | `minishlab/potion-retrieval-32M` | 512 | mit | | `mixedbread-ai/mxbai-embed-large-v1` | 1024 | apache-2.0 | | `nomic-ai/nomic-embed-text-v1` | 768 | apache-2.0 | | `nomic-ai/nomic-embed-text-v1.5` | 768 | apache-2.0 | | `nomic-ai/nomic-embed-text-v1.5-Q` | 768 | apache-2.0 | | `Qdrant/clip-ViT-B-32-text` | 512 | mit | | `Qdrant/constella-nano` | 1024 | mit | | `Qdrant/constella-zero` | 1024 | mit | | `Qdrant/stella-en-400M-v5-doc-onnx` | 1024 | mit | | `Qwen/Qwen3-Embedding-0.6B` | 1024 | apache-2.0 | | `Qwen/Qwen3-Embedding-0.6B-Q` | 1024 | apache-2.0 | | `sentence-transformers/all-MiniLM-L6-v2` | 384 | apache-2.0 | | `sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2` | 384 | apache-2.0 | | `sentence-transformers/paraphrase-multilingual-mpnet-base-v2` | 768 | apache-2.0 | | `snowflake/snowflake-arctic-embed-l` | 1024 | apache-2.0 | | `snowflake/snowflake-arctic-embed-m` | 768 | apache-2.0 | | `snowflake/snowflake-arctic-embed-m-long` | 768 | apache-2.0 | | `snowflake/snowflake-arctic-embed-s` | 384 | apache-2.0 | | `snowflake/snowflake-arctic-embed-xs` | 384 | apache-2.0 | | `thenlper/gte-base` | 768 | mit | | `thenlper/gte-large` | 1024 | mit | Follow [Generate Text Embeddings](https://qdrant.tech/documentation/fastembed/fastembed-quickstart/index.md) to load a model, or [Dense Embeddings with Qdrant](https://qdrant.tech/documentation/fastembed/fastembed-semantic-search/index.md) to index and search its vectors. ## Sparse Text Models These models produce sparse vectors. Configure a [sparse vector field](https://qdrant.tech/documentation/manage-data/vectors/index.md#sparse-vectors) in your collection. | Model | License | |---|---| | `opensearch-project/opensearch-neural-sparse-encoding-doc-v3-gte` | apache-2.0 | | `prithivida/Splade_PP_en_v1` | apache-2.0 | | `Qdrant/bm25` | apache-2.0 | | `Qdrant/bm42-all-minilm-l6-v2-attentions` | apache-2.0 | | `Qdrant/minicoil-v1` | apache-2.0 | See the [miniCOIL](https://qdrant.tech/documentation/fastembed/fastembed-minicoil/index.md) and [SPLADE](https://qdrant.tech/documentation/fastembed/fastembed-splade/index.md) guides for indexing and search examples. ## Text Multivector Models These models produce multiple vectors per text input. Dimensions refer to each vector in the output. | Model | Dimensions | License | |---|---|---| | `answerdotai/answerai-colbert-small-v1` | 96 | apache-2.0 | | `colbert-ir/colbertv2.0` | 128 | mit | | `jinaai/jina-colbert-v2` | 128 | cc-by-nc-4.0 | | `lightonai/LateOn` | 128 | apache-2.0 | | `lightonai/mLateOn` | 128 | apache-2.0 | Follow [ColBERT Multivector Embeddings](https://qdrant.tech/documentation/fastembed/fastembed-colbert/index.md) to configure a collection and search with multivectors. ## Image Models These models produce one dense vector per image. | Model | Dimensions | License | |---|---|---| | `google/siglip2-base-patch16-224` | 768 | apache-2.0 | | `jinaai/jina-clip-v1` | 768 | apache-2.0 | | `nomic-ai/nomic-embed-vision-v1.5` | 768 | apache-2.0 | | `nomic-ai/nomic-embed-vision-v1.5-Q` | 768 | apache-2.0 | | `Qdrant/clip-ViT-B-32-vision` | 512 | mit | | `Qdrant/resnet50-onnx` | 2048 | apache-2.0 | | `Qdrant/Unicom-ViT-B-16` | 768 | mit | | `Qdrant/Unicom-ViT-B-32` | 512 | mit | ## Image and Text Multivector Models These models support image and text inputs and produce multivectors. Dimensions refer to each vector in the output. | Model | Dimensions | License | |---|---|---| | `Qdrant/colmodernvbert` | 128 | mit | | `Qdrant/colpali-v1.3-fp16` | 128 | gemma | ## Store and Search the Vectors 1. [Create a Collection](https://qdrant.tech/documentation/manage-data/collections/index.md#create-a-collection): Match the model's dimensions and distance metric. 2. [Upload Vectors](https://qdrant.tech/documentation/manage-data/points/index.md#upload-points): Batch points with payloads for filtering. 3. [Search](https://qdrant.tech/documentation/search/search/index.md#query-api): Find matching points with the Query API. ## Next Steps - [Explore the FastEmbed guides](https://qdrant.tech/documentation/fastembed/index.md) for embedding generation and Qdrant integration. - [Rerank with FastEmbed](https://qdrant.tech/documentation/fastembed/fastembed-rerankers/index.md) to score retrieved results with a cross-encoder. - [Choose an inference approach](https://qdrant.tech/documentation/inference/index.md) to compare local embedding generation with Qdrant Cloud Inference. - [Filter search results](https://qdrant.tech/documentation/search/filtering/index.md) by payload fields. - [Combine dense and sparse vectors](https://qdrant.tech/documentation/search/hybrid-queries/index.md) for hybrid search. - [Reduce memory use with quantization](https://qdrant.tech/documentation/manage-data/quantization/index.md).