Nemotron 3.5 Embed 8B BF16
Oct 2026
Nemotron 3.5 Embed 8B BF16Compare inonViDoRe v3
$0.035per 1M tokens
60.2ViDoRe v3
Oct 2026released
4,096Dimensions
33K tokensMax input
en, es, fr +30Languages
Alternatives?
ModelScoreCostPriceWhat changes
CheaperNemotron 3.5 Embed 1B BF16Open-4%-57%$0.015
BetterEVIE 8BOpen+10%——Languages: not verifiedBetterEVIE 4.5BOpen+9%——Languages: not verifiedwBetterWebAI ColVec1.1 8BOpen+8%——Languages: not verified
Betterpplx Embed v2 Late 9BOpen+8%——Max input: 33K tokens → 4K tokens+1ModelScoreCostPriceWhat changes
Nemotron 3.5 Embed 1B BF16Open-4%-57%$0.015ModelScoreCostPriceWhat changes
EVIE 8BOpen+10%——Languages: not verifiedEVIE 4.5BOpen+9%——Languages: not verifiedwWebAI ColVec1.1 8BOpen+8%——Languages: not verifiedpplx Embed v2 Late 9BOpen+8%——Max input: 33K tokens → 4K tokens+1vVultronRetrieverPrime Qwen3.5 8BOpen+7%——Languages: loses zh, ja, ko +24ModelScoreCostPriceWhat changes
EVIE 8BOpen+10%——Languages: not verifiedEVIE 4.5BOpen+9%——Languages: not verifiedwWebAI ColVec1.1 8BOpen+8%——Languages: not verifiedpplx Embed v2 Late 9BOpen+8%——Max input: 33K tokens → 4K tokens+1vVultronRetrieverPrime Qwen3.5 8BOpen+7%——Languages: loses zh, ja, ko +24ModelScoreCostPriceWhat changes
Nemotron Colembed VL 8B v2Open+5%——Languages: not verified+1Nemotron Colembed VL 4B v2Open+2%——Languages: not verifiedLlama Nemotron Colembed VL 3B v2Opensame——Max input: 33K tokens → 8K tokens+1Nemotron 3.5 Embed 1B BF16Open-4%-57%$0.015Llama Nemoretriever Colembed 3B v1Open-5%——Max input: 33K tokens → 8K tokens+1Llama Nemoretriever Colembed 1B v1Open-8%——Max input: 33K tokens → 8K tokens+1Llama Nemotron Embed VL 1B v2Openn/a—$0.01Max input: 33K tokens → 10K tokens+1Omni Embed Nemotron 3BOpenn/a——Languages: loses es, fr, de +29All NVIDIA models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · compressed✓ 16 GB · 8-bit✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
Licenceopenmdw-1.1commercial use with conditions
Benchmarks
in Embeddings & RetrievalFAQ
How much does Nemotron 3.5 Embed 8B BF16 cost?›
$0.035 per 1M tokens.
Is Nemotron 3.5 Embed 8B BF16 free to use?›
There is no free API; the cheapest paid price is $0.035 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Nemotron 3.5 Embed 8B BF16, and what is it best at?›
#18 of 50 in Multimodal.
Is Nemotron 3.5 Embed 8B BF16 better than Ops Colqwen3 4B?›
On ViDoRe v3 · Mean over tasks · multilingual (6 languages), Nemotron 3.5 Embed 8B BF16 scores 60.2 vs 61.2 — about the same.
How can I use Nemotron 3.5 Embed 8B BF16?›
Through the NVIDIA API.
When was Nemotron 3.5 Embed 8B BF16 released? Is there a newer version?›
Released 2026-10-20. It is the current version in its line.
Is Nemotron 3.5 Embed 8B BF16 open source?›
Yes, the weights are open (openmdw-1.1 licence).
Can I run Nemotron 3.5 Embed 8B BF16 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is Nemotron 3.5 Embed 8B BF16?›
Nemotron 3.5 Embed 8B BF16 is an open-weights embeddings & retrieval model from NVIDIA, released Oct 2026.
What are the best alternatives to Nemotron 3.5 Embed 8B BF16?›
Nemotron 3.5 Embed 1B BF16 (cheaper), EVIE 8B (better).