5 AI models similar to text-embedding-3-small by OpenAI, ranked by performance.
| # | Model | Provider | Score | Rating | Context | Input $/1M | Compare |
|---|---|---|---|---|---|---|---|
| 1 | Voyage 3 Top-performing embedding model on MTEB with optimized retrieval quality for RAG applications and long-context support. | Voyage AI | 92.5 | ★ 4.6 | 32K | $0.06 | vs text-embedding-3-small |
| 2 | text-embedding-3-large OpenAI's most capable embedding model producing 3,072-dimensional vectors for search, RAG, classification, and clustering. | OpenAI | 91.0 | ★ 4.5 | 8K | $0.13 | vs text-embedding-3-small |
| 3 | Cohere Embed v3 Cohere's multilingual embedding model supporting 100+ languages with built-in search optimization. | Cohere | 90.0 | ★ 4.4 | 512 | $0.10 | vs text-embedding-3-small |
| 4 | BGE Large v1.5 BAAI's open-source embedding model competitive with commercial offerings, ideal for self-hosted deployments. | BAAI | 86.0 | ★ 4.3 | 512 | — | vs text-embedding-3-small |
| 5 | Nomic Embed Text v1.5 Fully open-source embedding model with Matryoshka dimension support and strong retrieval performance. | Nomic AI | 84.0 | ★ 4.2 | 8K | — | vs text-embedding-3-small |