Cohere has launched Rerank 5, a new retrieval model available in Pro and Fast tiers. Rerank 5 Pro reportedly outperforms existing models on benchmarks like nanoMTEB, ViDoRe V3, and BRIGHT, making it suitable for enterprise document retrieval. The Rerank 5 Fast tier offers significantly higher throughput and faster response times, outperforming smaller competitors and larger models alike. Additionally, Cohere introduced Context Focus, a feature designed to reduce token consumption by returning precise passages with relevance scores, potentially cutting costs by up to 41% while maintaining accuracy. AI
IMPACT Enhances retrieval accuracy and efficiency, potentially reducing costs for LLM applications by optimizing token consumption.
RANK_REASON New model release from a frontier AI lab (Cohere).
- Amazon SageMaker
- BRIGHT
- Cohere
- Cohere API
- Microsoft Foundry
- Model Vault
- nanoMTEB
- Qwen3-Reranker-8B
- Rerank 5
- Rerank 5 Fast
- Rerank 5 Pro
- ViDoRe V3
- Voyage AI Rerank 3 Lite
AI-generated summary · Google Gemini · from 5 sources. How we write summaries →