Llama Embed Nemotron 8B

Oct 2025
Llama Embed Nemotron 8B
Compare in
on
self-host onlyprice
79.6MTEB Scandinavian v1
Oct 2025released
4,096Dimensions
131K tokensMax input

Ranking?

Full ranking →
1Gemini Embedding 2 Preview95.82Jina Embeddings v5 Omni Small862Jina Embeddings v5 Text Small864Jina Embeddings v5 Omni Nano864Jina Embeddings v5 Text Nano86
···
13Harrier OSS v1 270M80.114Llama Embed Nemotron 8B79.615GritLM 7B76.9

Alternatives?

ModelScoreCostPriceWhat changes
BetterQwen 3 Embedding 8BunstatedOpen+6%—$0.01Max input: 131K tokens → 33K tokensOpenJina Embeddings v5 Text Smallunstated+8%—$0.05Max input: 131K tokens → 33K tokensBetterJina Embeddings v5 Text NanounstatedOpen+8%—$0.02Max input: 131K tokens → 8K tokensBetterQwen 3 Embedding 4BunstatedOpen+3%—$0.02Max input: 131K tokens → 33K tokensBetterHarrier OSS v1 27BOpen+7%——Max input: 131K tokens → 33K tokens+1

Run it yourself

estimate
Fits 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/llama-embed-nemotron-8b/blob/main/LICENSEcommercial use with conditions

Benchmarks

in Embeddings & Retrieval
MTEB Europe v1 · Bitext Mining · multilingual (28 languages)92.1#7
MTEB Europe v1 · Clustering · multilingual (29 languages)49.9#13
MTEB Europe v1 · Instruction Reranking · English10.8#3
MTEB Europe v1 · Reranking · multilingual (12 languages)62.5#12
MTEB Europe v1 · Multilabel Classification · multilingual (23 languages)22#13
MTEB French v1 · Pair Classification · French66.8#17
MTEB Indic v1 · Reranking · multilingual (2 languages)88.5#6
MTEB Indic v1 · Pair Classification · Hindi80.1#16
MTEB Indic v1 · STS · multilingual (13 languages)58.2#11
MTEB Indic v1 · Clustering · multilingual (21 languages)57#8
MTEB Korean v2 · Retrieval · multilingual (2 languages)56.3#28
MTEB Dutch v1 · Mean over tasks · Dutch66.2#2
MTEB Dutch v1 · Mean over task types · Dutch70.7#1
MTEB Dutch v1 · Classification · Dutch73.4#3
MTEB Dutch v1 · Clustering · Dutch50.8#1
MTEB Dutch v1 · Retrieval · Dutch67#4
MTEB Dutch v1 · STS · Dutch86.4#5
MTEB Dutch v1 · Multilabel Classification · Dutch41.5#4
MTEB Dutch v1 · Pair Classification · Dutch83.9#1
MTEB Dutch v1 · Reranking · Dutch91.7#14
MTEB Portuguese v1 · Multilabel Classification · Portuguese29#7
FollowIR · Mean over tasks · English10.8#3
FollowIR · Mean over task types · English10.8#3
FollowIR · Instruction Reranking · English10.8#3
MTEB Spanish v1 · Pair Classification · Spanish74.8#13
MTEB Thai v1 · Clustering · Thai63.4#10
MTEB Thai v1 · Pair Classification · Thai81.4#15
MTEB Scandinavian v1 · Bitext Mining · multilingual (3 languages)79.6#14
MTEB Russian v1.1 · Pair Classification · Russian67.2#10
MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages)69.5#5
MTEB multilingual v2 · Mean over task types · multilingual (1037 languages)61.1#5
MTEB multilingual v2 · Classification · multilingual (119 languages)73.2#6
MTEB multilingual v2 · Clustering · multilingual (208 languages)54.4#15
MTEB multilingual v2 · STS · multilingual (35 languages)79.4#5
MTEB multilingual v2 · Pair Classification · multilingual (24 languages)84#9
MTEB multilingual v2 · Retrieval · multilingual (120 languages)68.7#10
MTEB multilingual v2 · Bitext Mining · multilingual (1026 languages)81.7#5
MTEB multilingual v2 · Instruction Reranking · English10.8#3
MTEB multilingual v2 · Multilabel Classification · multilingual (25 languages)29.9#10
MTEB multilingual v2 · Reranking · multilingual (22 languages)67.8#7

FAQ

How much does Llama Embed Nemotron 8B cost?›

There is no per-unit API price: self-host only.

Is Llama Embed Nemotron 8B free to use?›

There is no free API. The weights are open, so you can run it yourself.

How good is Llama Embed Nemotron 8B, and what is it best at?›

#11 of 205 in Embeddings.

Is Llama Embed Nemotron 8B better than Harrier OSS v1 270M?›

On MTEB Scandinavian v1 · Bitext Mining · multilingual (3 languages), Llama Embed Nemotron 8B scores 79.6 vs 80.1 — about the same.

When was Llama Embed Nemotron 8B released? Is there a newer version?›

Released 2025-10-23. It is the current version in its line.

Is Llama Embed Nemotron 8B open source?›

Yes, the weights are open (https://huggingface.co/nvidia/llama-embed-nemotron-8b/blob/main/LICENSE licence).

Can I run Llama Embed Nemotron 8B locally?›

The weights are open and can be self-hosted; hardware needs are not confirmed yet.

What is Llama Embed Nemotron 8B?›

Llama Embed Nemotron 8B is an open-weights embeddings & retrieval model from NVIDIA, released Oct 2025.

What are the best alternatives to Llama Embed Nemotron 8B?›

Qwen 3 Embedding 8B (better), Jina Embeddings v5 Text Small (open weights).