Voyage Rerank 2.5
What ChatGPT, Claude, Gemini & Grok actually say · July 2026
The verdict
Voyage Rerank 2.5 appears in 1 AI-ranked category — best position #1 for reranking model api.
Excellent general-purpose quality with instruction following, multilingual retrieval, a 32K context window, and support for up to 1,000 candidates; arguably the better choice than Cohere for long documents.
Claude Benchmark leader on hard retrieval evals with instruction-following reranking and long context windows that reduce chunking gymnastics; aggressive pricing and a lite variant for latency-sensitive paths; MongoDB acquisition gives it real staying power; near-tie with Cohere — pick it when raw quality per dollar matters more than marketplace breadth
Grok Highest relevance lift in many independent RAG tests; longest context (16k tokens); excels on complex/domain-heavy retrieval (finance, legal).
Gemini Delivers exceptional retrieval quality and domain-specific search performance, backed by strong enterprise integrations (MongoDB and Snowflake) and a low-latency Lite variant.
Where Voyage Rerank 2.5 falls short, per the models
- GPT Proprietary hosted model with less enterprise deployment flexibility than Cohere.
- Claude Smaller ecosystem and fewer cloud-marketplace/VPC deployment options than Cohere, and the MongoDB tie-up is steering the roadmap toward Atlas integration
- Gemini Closed-weights only, forcing complete reliance on the Voyage API and showing optimal performance only when paired with Voyage's own embeddings.
- Grok Cut p99 latency and per-query costs to win high-volume real-time workloads.
Top alternatives per the models: Cohere Rerank 4 · Jina Reranker v3 · ZeroEntropy zerank-2 · Cohere Rerank 3.5
Rankings are computed from what the models answer, re-polled continuously · raw reasoning shown verbatim · methodology