Text embeddings

multilingual-e5-large-instruct

Microsoft (intfloat) · open weights

$0.010 per 1M input tokens, cheapest of 1 seller (DeepInfra)

Dominated: EmbeddingGemma 300M is at least as good and at least as cheap on the same benchmark task, with a strict advantage on one axis. As of .

What was measured

BoardMean (Task)Mean (TaskType)TasksRevision
MTEB(eng, v2)65.5661.2441 of 41baa7be480a7de1539afce709c8f13f833a510e0a

By task type: Classification 75.71 · Clustering 49.89 · PairClassification 86.24 · Reranking 48.74 · Retrieval 53.47 · STS 84.72 · Summarization 29.89

What sellers charge

SellerAs the seller states itConditionsNormalised per 1M input tokensSource
DeepInfra1e-8 per_token—$0.010 per token × 1,000,000api.deepinfra.com read 2026-11-05 · feed

Provenance

Read the full source record

MTEB results, https://github.com/embeddings-benchmark/results, released under CC0-1.0; MTEB(eng, v2) means computed from the per-task files as described in data/modalities/embeddings/benchmark.json.

Evidence & Ask