Embedding Models

AWS's embedding model via Bedrock. 8K token input, 1024 dimensions. Native integration with AWS services.

BAAI General Embedding — a family of open-source embedding models from Beijing Academy of AI. BGE-M3 supports 100+ langu...

Cohere Embed converts text and images into vector embeddings for semantic search, RAG, and agentic applications. Support...

Cross-encoder reranking model. Improves retrieval quality by reranking search results for relevance.

Late-interaction retrieval model by Stanford. Token-level matching for high-quality passage retrieval.

Microsoft's embedding models. Instruct-tuned variants for task-specific retrieval with natural language instructions.

Lightweight Python library for embedding generation. ONNX-based, runs on CPU without PyTorch dependency.

Google's latest embedding model via Vertex AI. 768 dimensions, optimized for retrieval and classification.

Google's text-embedding-004 and multimodal-embedding models via Vertex AI. Supports text and image embeddings with up to...

Open-source 7B-parameter text embedding model built on Qwen2 LLM for generating high-quality vector representations. Sup...

Embedding API from Jina AI featuring jina-embeddings-v3 with 8192-token context, multilingual support (89 languages), an...

Cross-encoder reranking model by Jina AI. Improve retrieval quality with semantic relevance scoring.

Mistral's embedding model. 1024 dimensions, 8K context. Competitive pricing via Mistral API.

High-quality embedding models with Matryoshka support. Reduce dimensions without retraining for cost savings.

Fully open-source, auditable embedding model (nomic-embed-text-v1.5) with 8192-token context window. Available via Nomic...

NVIDIA's embedding model. Top MTEB scores with strong retrieval performance across domains.

Generates high-dimensional vector embeddings from text using OpenAI's latest embedding model. Supports flexible dimensio...

OpenAI's highest-quality text embedding model. 3072 dimensions with Matryoshka representation learning — truncate to any...

Open-source Python library (by UKP Lab / Hugging Face) for computing dense vector representations of sentences and parag...

Solar Embeddings is an embedding API by Upstage that converts text into dense vector representations. It supports multil...

Sparse embedding model for information retrieval. Combines neural and lexical matching for hybrid search.

Embedding specialist acquired by Anthropic in 2024. Offers voyage-3, voyage-3-lite, and domain-specific models (voyage-c...

Code-specialized embedding model by Voyage AI. Best-in-class for code search and retrieval tasks.