Embedding Models
AI models for text embeddings and semantic search
Embedding
- Verified price— Discount
- —
- Official price
- Input$0.0184/1MOutput$0.00/1M
Supports prompt caching
- Verified price— Discount
- —
- Official price
- Input$0.15/1MOutput$0.00/1M
- Verified price-30% Discount
- Input$0.091/1MOutput$0.00/1M
- Official price
- Input$0.13/1MOutput$0.00/1M
- Verified price-30% Discount
- Input$0.014/1MOutput$0.00/1M
- Official price
- Input$0.02/1MOutput$0.00/1M
All model series
6 series
Gemini Embedding
Embedding
Auto from $0.075 · per 1M tokens
2 modelsOpen series
Doubao
ChatEmbedding+3
Auto from $0.10 · per 1M tokens
10 modelsOpen series
Voyage Embedding
Embedding
Auto from $0.02 · per 1M tokens
10 modelsOpen series
Qwen Embedding
EmbeddingVision
Auto from $0.0184 · per 1M tokens
5 modelsOpen series
Mistral
ChatEmbedding+1
Auto from $0.10 · per 1M tokens
3 modelsOpen series
Text Embedding
Embedding
Auto from $0.014 · per 1M tokens
3 modelsOpen series
Choose embeddings for retrieval
An embedding model maps content to vectors for search and related tasks. Evaluate retrieval quality on your own queries and documents before optimizing cost or vector size.
Selection signals
- Check input limits, supported languages, and output dimensions against your vector index.
- Compare ingestion and repeated-query costs using the same document chunking strategy.
- Keep document and query embeddings compatible. Changing the embedding model can require re-embedding the index.