Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
Scroll to compare models →
| 3 models | intfloatmultilingual-e5-large-instructRemove × | BAAIbge-m3Remove × | googleGoogle: Gemini Embedding 2Remove × |
|---|---|---|---|
| What’s different | The Multilingual-E5 models, initialized from XLM-RoBERTa, support up to 512 tokens per input — any longer text will be silently truncated. To ensure optimal performance, always prefix inputs with “query:”… | BGE-M3 is a versatile text embedding model that supports multi-functionality, multi-linguality, and multi-granularity, allowing it to perform dense retrieval, multi-vector retrieval, and sparse retrieval in… | Gemini Embedding 2 is Google's first multimodal embedding model. We currently support mapping text and images into a unified vector space for semantic search and retrieval-augmented generation (RAG). It… |
| Date | Apr 17, 2025Listed | May 27, 2024Listed | May 20, 2026Listed |
API prices USD · provider configuration shown below | |||
| Listed rate | See provider quotes | $0.01 in / $0.00 out/ 1M tokensOpenRouter · all routes ↗ | $0.20 in / $0.00 out/ 1M tokensOpenRouter · all routes ↗ |
| Provider endpoints | 0 endpoints → | 1 endpoint → | 1 endpoint → |
Specifications | |||
| Context window | 512 | 8.2K | 8.2K |
| Max output | — | 7.4K | 7.4K |
| Parameters | — | — | — |
| Image input | Not reported | Not reported | Supported |
| Audio input | Not reported | Not reported | Supported |
| Video input | Not reported | Not reported | Supported |
| Open weights | Not reported | Supported | Not reported |