Compare Embedding Models

Side-by-side comparison of embedding model specifications

	Qwen3-Embedding-8B▾ Jina Embeddings v4 Jina Embeddings v3 ✓ Qwen3-Embedding-8B Gemini Embedding 001 mxbai-embed-large-v1 Cohere Embed v4 Cohere Embed English v3 Cohere Embed Multilingual v3 Voyage 3 Large GTE-Qwen2-7B-instruct Linq-Embed-Mistral LFM2-ColBERT-350M Jina Embeddings v5 Omni Small Jina Embeddings v5 Omni Nano Jina Embeddings v5 Omni Nano Text-Matching Jina Embeddings v5 Text Small Jina Embeddings v5 Text Nano Jina Embeddings v5 Text Small Text-Matching Qwen3-Embedding-8BQwen	Voyage 3 Large▾ Jina Embeddings v4 Jina Embeddings v3 Qwen3-Embedding-8B Gemini Embedding 001 mxbai-embed-large-v1 Cohere Embed v4 Cohere Embed English v3 Cohere Embed Multilingual v3 ✓ Voyage 3 Large GTE-Qwen2-7B-instruct Linq-Embed-Mistral LFM2-ColBERT-350M Jina Embeddings v5 Omni Small Jina Embeddings v5 Omni Nano Jina Embeddings v5 Omni Nano Text-Matching Jina Embeddings v5 Text Small Jina Embeddings v5 Text Nano Jina Embeddings v5 Text Small Text-Matching Voyage 3 LargeVoyage AI	Jina Embeddings v5 Omni Small▾ Jina Embeddings v4 Jina Embeddings v3 Qwen3-Embedding-8B Gemini Embedding 001 mxbai-embed-large-v1 Cohere Embed v4 Cohere Embed English v3 Cohere Embed Multilingual v3 Voyage 3 Large GTE-Qwen2-7B-instruct Linq-Embed-Mistral LFM2-ColBERT-350M ✓ Jina Embeddings v5 Omni Small Jina Embeddings v5 Omni Nano Jina Embeddings v5 Omni Nano Text-Matching Jina Embeddings v5 Text Small Jina Embeddings v5 Text Nano Jina Embeddings v5 Text Small Text-Matching Jina Embeddings v5 Omni SmallJina AI	Linq-Embed-Mistral▾ Jina Embeddings v4 Jina Embeddings v3 Qwen3-Embedding-8B Gemini Embedding 001 mxbai-embed-large-v1 Cohere Embed v4 Cohere Embed English v3 Cohere Embed Multilingual v3 Voyage 3 Large GTE-Qwen2-7B-instruct ✓ Linq-Embed-Mistral LFM2-ColBERT-350M Jina Embeddings v5 Omni Small Jina Embeddings v5 Omni Nano Jina Embeddings v5 Omni Nano Text-Matching Jina Embeddings v5 Text Small Jina Embeddings v5 Text Nano Jina Embeddings v5 Text Small Text-Matching Linq-Embed-MistralLinq AI
SPECIFICATIONS
Max dimension	4K	2K	1K	4K
Max tokens	33K	32K	33K	4K
Parameters	8B	—	1.74B	7B
License	Apache 2.0	Proprietary	CC BY-NC 4.0	Apache 2.0
EMBEDDING TYPES
Types	Dense	Dense	Dense	Dense
MATRYOSHKA DIMENSIONS
Available sizes	3264128256512102420484096	25651210242048	32641282565127681024	—
INPUT MODALITIES
Text	✓	✓	✓	✓
Image	—	—	✓	—
PDF	—	—	✓	—
OUTPUT TYPES
Single vector	✓	✓	—	✓
Multi vector	—	—	—	—
LANGUAGE SUPPORT
Languages	🌍 100+	🌍 Multi	—	🌍 Multi
PRICING
Per 1M tokens	—	—	—	—
BENCHMARKS
CMTEB CHINESE	73.84	—	—	—
MTEB AVG	—	—	—	68.20
MTEB EN V2	75.22	—	—	—
MTEB MULTILINGUAL	70.58	—	—	—
MTEB RETRIEVAL	—	—	—	60.20