Nemotron Super 49B
A mid-sized open-weight model that keeps much of Ultra's reasoning style while fitting comfortably on a single high-end GPU node.
At a glance
Released
Mar 5, 2026
Provider
NVIDIA
Context window
256,000 tokens
Max output
16,000 tokens
Modalities
Text → text
Knowledge cutoff
December 2025
API model ID
nemotron-super-49bLicense
Open weights (NVIDIA Open Model License)
Pricing
Standard
Input$0.80 / 1M
Output$3.20 / 1M
* Rates are from a third-party aggregator — verify before relying on them. A typical 1,500-token prompt with a 650-token answer costs ≈ $0.003 on the standard tier. Run your own prompt through the calculator →
Best use cases
Cost-aware reasoning
Delivers solid multi-step reasoning without the Ultra footprint.
Enterprise RAG
Grounds answers in private knowledge bases on-prem via NIM.
Agent backends
Serves as a dependable, self-hostable engine for tool-using agents.
Strengths & limitations
Strengths
- Good reasoning per dollar
- Single-node friendly
- Open weights
Limitations
- Below Ultra on hardest problems
- Text only
How to test it
- No-code: open build.nvidia.com (NVIDIA NIM) and run a prompt in the playground.
- API: create a key in build.nvidia.com (NVIDIA NIM) and call the API (example below).
curl https://integrate.api.nvidia.com/v1/chat/completions \
# OpenAI-compatible endpoint
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "nemotron-super-49b",
"messages": [{"role": "user", "content": "Explain tokenization in one paragraph."}]
}'