Skip to content
Local AI Wizard
Your machine
Theme

Llama Nemotron Embed 1B v2

NVIDIA · nvidia/llama-nemotron-embed-1b-v2

Weights license: Publisher terms: NVIDIA Open Model License Agreement (also Llama 3.2 Community Model License Agreement) · code and components may use other terms

MTEB (Multilingual, v2)No sourced MTEB resultNo sourced MTEB result
Memory needed~4 GB8-bit planning estimate · not measured
Parameters1.24 BSee technical details for the basis
Dimensions · context2,0488,192 tokens · Matryoshka 1024/768/512/384

Quality

Not in the MTEB ranking

No sourced MTEB (Multilingual, v2) mean(task) result for this model in the MTEB results dataset or the model card; none is inferred.

Reported by the publisher

The card has 4 evaluation tables, but none of their columns has been matched to this checkpoint yet (read 2026-10-25); no value is copied from another column. This stays an open gap. tables on the card, column not matched yet · read 2026-10-25

Quality × memory

No comparable quality × memory pair

Embedding memory estimates cluster at a few GB, so a quality × memory chart would not separate these models; MTEB is ranked above instead.

Runs on

Fast memory holds it 119

The planning estimate fits the memory the model can use quickly: graphics card VRAM, or unified memory within the macOS GPU share.

Show all 119

Hardware profiles grouped by the embedding planning estimate (8-bit weights, 20% allowance, 2 GB reserve). Not a measured fit.

Consider it for

Consider Llama Nemotron Embed 1B v2 for retrieval and semantic search: it turns text into 2,048-dimension vectors.

Check first

The publisher states no memory figure; the hardware list uses a ~4 GB planning estimate, a ceiling and not a measurement. See the prompt format under How to run. PT-BR is confirmed by a source.

Measured speed

No published measurement for this checkpoint

Throughput (texts per second) is shown only when a source publishes it with hardware, build and setup. It is never estimated.

How to run

The planning estimate uses 8-bit weights, 20% working allowance and a 2 GB reserve; it is a ceiling, not a measurement.

Runs in
Sentence Transformers, vLLM, NVIDIA NIMsource
Query/document prompts
Prefix queries with 'query:' and documents with 'passage:' (the Sentence Transformers path uses encode_query / encode_document).
Languages · PT-BR
Evaluated on 26 languages, Portuguese includedPT-BR confirmed by a source

Documented ways to run it

Copied as printed in the linked source (read 2026-10-18); not run by this site.

Setup stated first by the source · Sentence Transformers
pip install transformers sentence-transformers
Sentence Transformers · model card · **Sentence Transformers Usage**
from sentence_transformers import SentenceTransformer

model = SentenceTransformer("nvidia/llama-nemotron-embed-1b-v2", trust_remote_code=True)

queries = [
    "how much protein should a female eat",
    "summit define",
]
documents = [
    "As a general guideline, the CDC's average requirement of protein for women ages 19 to 70 is 46 grams per day. But, as you can see from this chart, you'll need to increase that if you're expecting or training for a marathon. Check out the chart below to see how much protein you should be eating each day.",
    "Definition of summit for English Language Learners. : 1  the highest point of a mountain : the top of a mountain. : 2  the highest level. : 3  a meeting or series of meetings between the leaders of two or more governments."
]

query_embeddings = model.encode_query(queries, convert_to_tensor=True)
document_embeddings = model.encode_document(documents, convert_to_tensor=True)

# Compute similarity scores
scores = model.similarity(query_embeddings, document_embeddings)
"""
tensor([[ 0.5968, -0.0454],
        [-0.0336,  0.4613]], device='cuda:0')
"""
vLLM · model card · vLLM Usage
vllm serve \
    nvidia/llama-nemotron-embed-1b-v2 \
    --trust-remote-code
Setup stated first by the source · Transformers
pip install transformers
Transformers · model card · **Transformers Usage**
import torch
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel


def average_pool(last_hidden_states, attention_mask):
    """Average pooling with attention mask."""
    last_hidden_states_masked = last_hidden_states.masked_fill(~attention_mask[..., None].bool(), 0.0)
    embedding = last_hidden_states_masked.sum(dim=1) / attention_mask.sum(dim=1)[..., None]
    embedding = F.normalize(embedding, dim=-1)
    return embedding


tokenizer = AutoTokenizer.from_pretrained("nvidia/llama-nemotron-embed-1b-v2")
model = AutoModel.from_pretrained("nvidia/llama-nemotron-embed-1b-v2", trust_remote_code=True)
model = model.to("cuda:0")
model.eval()
query_prefix = "query:"
document_prefix = "passage:"


queries = [
    "how much protein should a female eat",
    "summit define",
]
documents = [
    "As a general guideline, the CDC's average requirement of protein for women ages 19 to 70 is 46 grams per day. But, as you can see from this chart, you'll need to increase that if you're expecting or training for a marathon. Check out the chart below to see how much protein you should be eating each day.",
    "Definition of summit for English Language Learners. : 1  the highest point of a mountain : the top of a mountain. : 2  the highest level. : 3  a meeting or series of meetings between the leaders of two or more governments."
]
queries = [f"{query_prefix} {query}" for query in queries]
documents = [f"{document_prefix} {document}" for document in documents]


batch_queries = tokenizer(queries, padding=True, truncation=True, return_tensors='pt').to("cuda:0")
batch_documents = tokenizer(documents, padding=True, truncation=True, return_tensors='pt').to("cuda:0")

with torch.no_grad():
    outputs_queries = model(**batch_queries)
    outputs_documents = model(**batch_documents)

# Average Pooling
embeddings_queries = average_pool(outputs_queries.last_hidden_state, batch_queries["attention_mask"])
print("Query embeddings:")
print(embeddings_queries)
print(embeddings_queries.shape)
#torch.Size([2, 2048])

embeddings_documents = average_pool(outputs_documents.last_hidden_state, batch_documents["attention_mask"])
print("\nDocument embeddings:")
print(embeddings_documents)
print(embeddings_documents.shape)
#torch.Size([2, 2048])

# Compute similarity scores
scores = (embeddings_queries @ embeddings_documents.T)
print("\nSimilarity scores:")
print(scores.tolist())

#Similarity scores:
#[[0.5968121290206909, -0.04534469544887543], [-0.03361201286315918, 0.46140915155410767]]

Choose my computer

Technical details

What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.

Task
embeddings for search (feature extraction)
Library in source
Sentence Transformers, vLLM, NVIDIA NIM
Weight license
Publisher terms: NVIDIA Open Model License Agreement (also Llama 3.2 Community Model License Agreement)
Commercial use
Allowed with conditions Both licenses apply: the NVIDIA Open Model License states the model is commercially usable; the Llama 3.2 terms add that over 700M monthly users needs a Meta license and "Built with Llama" must be shown · Llama 3.2 Community License (meta-llama/llama-models) · read 2026-10-18license text
Parameter basis
Hub safetensors total (1,235,814,400 parameters, read 2026-10-09 from the Hub API).
Released
Not published by the publisher. Full card read (overview through safety subcards): no Release Date field, unlike the Nemotron-3-Embed card.Checked 2026-10-13.source checked

Sources

Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.

Curated facts checked: 2026-10-09. Each value cites its source (the publisher's card or the MTEB results dataset); estimates are labelled as estimates, and a value no source states stays empty. A repository commit is not a model release date.