Cohere launches Embed 5: preference-aware retrieval, stronger multilingual, big throughput gains
cohere · x · 2026-10-01
Cohere's embeddings team detailed Embed 5's key innovations: it retrieves not just relevant results but the most human-preferred documents via new training and evaluation techniques, with a strong multilingual focus, massive throughput gains that scale to any data size, and a privacy-first stance that keeps customer data private.
Related event: Cohere Launches Embed 5 Enterprise Embedding Models in Pro and Fast Tiers(6 posts)→
More from Models
- OpenAI makes Luna free for subscribers; dev says gap to bigger models is now narrow — dosco · 2026-10-01
- 8B model trained on distributed gaming GPUs for $6,500 runs on phone CPU at ~60 tok/s — markjeffrey · 2026-10-01
- Praxis-1: An Open-Weight World Action Model Launches — c_valenzuelab · 2026-10-01
- Codex Pro's new credit pool undercuts value: code reviewer hits limits with 28% weekly usage left — EffectiveWebDev404 · 2026-10-01
- Ling-3.1-flash surfaces: ~560B total params, ~25B active per token, up to 1M context — Muhlwa_Sholanke · 2026-10-01
- From Opus 5 Fumble to Opus 5.5 Praise: What Changed at Anthropic? — rudrank · 2026-10-01