Cohere launches Embed 5 Pro and Fast for enterprise search and agents
Cohere released Embed 5 on September 30, a multimodal embedding model family split into two tiers: Embed 5 Pro (optimized for retrieval quality) and Embed 5 Fast (optimized for speed and cost). Both support text, images, and fused text-image inputs, cover 128K tokens, and 100+ languages. A key feature is their shared embedding space—users can index with Pro and query with Fast without…
Key points
- Embed 5 Pro scores 85.8 on ViDoRe V3, outperforming Voyage 4 Large and Gemini Embedding 2
- Embed 5 Fast costs $0.08 per 1M text tokens and processes 377.3 docs/sec, 2.4x faster than Pro
- Pro and Fast share one embedding space: index with Pro, query with Fast for speed without re-indexing
The models are now generally available via Cohere’s API, Model Vault, Microsoft Foundry, and Amazon SageMaker. Benchmarks show Embed 5 Pro scoring 85.8 on ViDoRe V3, ahead of Voyage 4 Large (83.7) and Gemini Embedding 2 (83.2). Embed 5 Fast costs $0.08 per 1M text tokens, while Pro costs $0.12, with image inputs priced at $0.40 per 1M tokens. Fast processes 377.3 documents per second, nearly 2.4x faster than Pro. Storage efficiency is also highlighted: 256-dim binary vectors use just 32 bytes, a 256x reduction compared to 2048-dim float32 vectors.
Model page: Embed 5 Pro →
The story so far
2 episodes →- Cohere launches Embed 5 Pro and Fast for enterprise search and agentsthis story
Cohere Releases Embed 5: How It Compares to Voyage 4 Large, Gemini Embedding 2, and OpenAI
MarkTechPost · 1 October 2026
Loading the full article…
This text was published by MarkTechPost and written by Sana Hassan. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Enterprise & Industry
All →- Cloudera’s Brunnick outlines hybrid AI platform strategy for enterprises · 1 src
- OpenAI pauses AI training again after agents escaped sandbox · 1 src
- AWS shows how contextual bandits lift conversions in acquisition funnels · 1 src
- Unily CEO explains why AI adoption depends on trust, not just tech · 1 src
- AWS details multi-environment setup for Claude Platform on AWS · 1 src
Comments
via GitHub Discussions