Build and deploy clinical AI agents with NVIDIA Nemotron and Nebius Token Factory
Nemotron gives you the model. Token Factory ships it. You own the result.
Open models with the infrastructure to deploy them — built for the complexity of healthcare.
Why digital health teams use Token Factory
Own your model foundation
Healthcare moves too fast for a one-size-fits-all approach. Nemotron open weights mean you can fine-tune on your own clinical datasets, adapt to your patient population, and build a model that compounds in value over time.
Built for regulated industries like Healthcare
From a startup scaling across markets to a health system with strict data residency requirements, Nemotron deploys on the infrastructure you own. No shared cloud dependencies, no data leaving your environment.
From evaluation to production, with urgency
Nebius Token Factory removes the infrastructure bottleneck — serverless fine-tuning, dedicated endpoints, and per-token pricing so you can move from model selection to shipped agent without standing up your own compute.
Nemotron family models at Token Factory
NVIDIA NemotronTM is a family of open models, weights, and libraries purpose-built for customization, giving digital health companies the foundation to deploy specialized clinical agents across ambient care, decision support, and patient outcome modeling.
Nemotron 3 Nano 30b
Compact MoE model optimized for efficient reasoning, chat, and coding with strong multilingual support and long-context RAG/agent workflows.
Nemotron 3 Nano Omni
The most open, efficient, and accurate omni-modal reasoning model for agentic AI.
Nemotron 3 Super 120b
Hybrid MoE model optimized for efficient multi-agent AI and complex reasoning tasks.
Nemotron 3 Ultra 550b
Frontier hybrid MoE model optimized for long-running autonomous agents, deep research, and high-throughput workflows.
Why use Nemotron
Nemotron on Token Factory can improve output quality and performance while reducing inference costs.
reduced inference spend
reduced latency
improved throughput
Choose your path for a quick start
Playground
Playground
Experiment, compare models and tune prompts side-by-side.
OpenAI-compatible API
Move your endpoint to Token Factory and switch models in seconds.
import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.tokenfactory.us-central1.nebius.com/v1/",
api_key=os.environ.get("NEBIUS_API_KEY"))
response = client.chat.completions.create(
model="nvidia/nemotron-3-super-120b-a12b",
messages=[
{"role": "system", "content": "You are helpful assistant"},
{"role": "user", "content": [{"type": "text", "text": "Hello"}]}])
print(response.to_json())
Start building today
