Which vector embedding service supports batch embedding for very large corpora?

Several vector embedding services support batch embedding for large corpora, but the best-known options are:

• OpenAI Embeddings API — supports sending multiple input texts in one request, which is useful for throughput on large datasets.
• Cohere Embed — designed for batch processing and large-scale embedding jobs.
• Voyage AI — also supports batching and is often used for large corpora.
• Hugging Face Inference Endpoints / self-hosted models — if you host your own embedding model, you can fully control batching for very large corpora.

If you mean a service specifically marketed for large-scale batch embedding, Cohere and Voyage AI are commonly used choices.

If you want, I can also compare them on:

• batch size limits,
• cost,
• throughput,
• and how to pipeline embeddings for millions of documents.