Omni Embed Nemotron 3B
Oct 2025
Omni Embed Nemotron 3BCompare inon
Omni Embed Nemotron 3B measured on 20 of 203 Multimodal benchmarks20
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
10.9MVEB text video v1
Oct 2025released
2,048Dimensions
33K tokensMax input
enLanguages
Alternatives?
ModelScoreCostPriceWhat changes
BetterJina Embeddings v5 Omni SmallunstatedOpen+35%—$0.05Languages: not verified
OpenQwen 3 VL Embedding 8B+70%—$0.09Languages: not verified+1BetterQwen 3 VL Embedding 2BOpen+62%——Languages: not verifiedLBetterLCO Embedding Omni 7BunstatedOpen+58%——Languages: not verified+1LBetterLCO Embedding Omni 3BunstatedOpen+53%——Languages: not verifiedModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
Jina Embeddings v5 Omni SmallunstatedOpen+35%—$0.05Languages: not verifiedQwen 3 VL Embedding 8BOpen+70%—$0.09Languages: not verified+1Qwen 3 VL Embedding 2BOpen+62%——Languages: not verifiedLLCO Embedding Omni 7BunstatedOpen+58%——Languages: not verified+1LLCO Embedding Omni 3BunstatedOpen+53%——Languages: not verifiedModelScoreCostPriceWhat changes
Qwen 3 VL Embedding 8BOpen+70%—$0.09Languages: not verified+1Qwen 3 VL Embedding 2BOpen+62%——Languages: not verifiedLLCO Embedding Omni 7BunstatedOpen+58%——Languages: not verified+1LLCO Embedding Omni 3BunstatedOpen+53%——Languages: not verifiedHE5 Omni 7BunstatedOpen+51%——Languages: not verified+1ModelScoreCostPriceWhat changes
Nemotron 3.5 Embed 1B BF16Openn/a—$0.015Nemotron 3.5 Embed 8B BF16Openn/a—$0.035Memory (4-bit): 2.8 GB → 5 GBLlama Nemotron Colembed VL 3B v2Openn/a——Max input: 33K tokens → 8K tokens+1Nemotron Colembed VL 4B v2Openn/a——Languages: not verified+1Nemotron Colembed VL 8B v2Openn/a——Languages: not verified+1Llama Nemotron Embed VL 1B v2Openn/a—$0.01Max input: 33K tokens → 10K tokens+1Llama Nemoretriever Colembed 1B v1Openn/a——Max input: 33K tokens → 8K tokens+1Llama Nemoretriever Colembed 3B v1Openn/a——Max input: 33K tokens → 8K tokens+1All NVIDIA models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · 8-bit✓ 16 GB · full quality✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
Licencehttps://huggingface.co/nvidia/omni-embed-nemotron-3b/blob/main/LICENSEcommercial use with conditions
Benchmarks
in Embeddings & RetrievalFAQ
How much does Omni Embed Nemotron 3B cost?›
There is no per-unit API price: self-host only.
Is Omni Embed Nemotron 3B free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Omni Embed Nemotron 3B, and what is it best at?›
#11 of 27 in Multimodal.
Is Omni Embed Nemotron 3B better than Xclip Base Patch32?›
On MVEB text video v1 · Mean over tasks · multilingual (16 languages), Omni Embed Nemotron 3B scores 35.8 vs 35.9 — about the same.
When was Omni Embed Nemotron 3B released? Is there a newer version?›
Released 2025-10-01. It is the current version in its line.
Is Omni Embed Nemotron 3B open source?›
Yes, the weights are open (https://huggingface.co/nvidia/omni-embed-nemotron-3b/blob/main/LICENSE licence).
Can I run Omni Embed Nemotron 3B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is Omni Embed Nemotron 3B?›
Omni Embed Nemotron 3B is an open-weights embeddings & retrieval model from NVIDIA, released Oct 2025.
What are the best alternatives to Omni Embed Nemotron 3B?›
Jina Embeddings v5 Omni Small (better), Qwen 3 VL Embedding 8B (open weights).