What is MTEB and why does it matter?

MTEB is the Massive Text Embedding Benchmark, a public suite that scores embedding models across retrieval, classification, clustering, and more. A higher average is a useful signal of general quality, but it is not a substitute for testing on your own domain data.

Do more dimensions always mean better embeddings?

No. Higher-dimensional vectors can capture more nuance but cost more to store and search and are slower at scale. Many modern models support dimension reduction (Matryoshka) so you can trade a little quality for much cheaper storage and faster search.

How do I read the price column?

Prices are shown per million input tokens, the standard unit providers bill on. To estimate cost, multiply your total tokens to embed by the price and divide by one million. Open-source models show as self-hosted, where you pay infrastructure rather than per token.

Which model is best for retrieval-augmented generation?

For RAG, prioritize retrieval performance and max input tokens so you can embed larger chunks. A high MTEB retrieval score and strong multilingual support matter most; the table lets you filter to strong general models and compare their context limits.

Are these numbers exact and current?

They are representative figures to guide a shortlist, and providers update models and pricing frequently. Always confirm the latest dimensions, limits, and prices on the provider's own documentation before committing.

Embedding Model Comparison Table

Embedding model comparison table

Choosing a text embedding model means balancing quality, vector size, context limit, and cost. This filterable table puts the popular options — from OpenAI, Cohere, Voyage AI, Jina, and open-source providers — side by side so you can build a shortlist by dimensions, max tokens, price, and benchmark score, then validate the finalists on your own data.

How it works

Each row lists a model with its native embedding dimensions, maximum input tokens, approximate price per million tokens, a representative MTEB-style benchmark score, and notes on multilingual support and dimension flexibility. You pick a task type and optionally filter by cost or dimension size; the table re-ranks and highlights the models best suited to your use. All filtering runs locally in your browser — no data is sent anywhere.

Notes and caveats

Benchmark scores guide, not decide. A high MTEB average suggests general quality, but domain-specific performance can differ — always test on your data.
Watch the dimension cost. Larger vectors cost more to store and search; Matryoshka-capable models let you shorten vectors with minimal quality loss.
Mind the context limit. For RAG with large chunks, max input tokens can matter more than a small benchmark difference.
Verify pricing. Providers change models and prices often; confirm current figures on the official docs before you commit.