Nvidia
NVIDIA: Nemotron 3 Ultra
nvidia/nemotron-3-ultra-550b-a55b
Score70
Add to CompareInput / M
Output / M
Context
Max Output
16K
Usage
Unranked
Quality
60.0 / 100
Providers
3
Best Uptime 30m
100.00%
Released
Jun 4, 2026
Modalities
text
About
NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...
Capabilities
Vision
Tools
Reasoning
Streaming
JSON Mode
Providers · 3
Per-provider pricing, uptime
| Provider | Input/M | Output/M | Context | Max Out | Quant | Uptime 30m |
|---|---|---|---|---|---|---|
| NebiusRecommended Nebius | nvidia/nemotron-3-ultra-550b-a55b-20260604 | $1.00 | $3.00 | 8K | — | fp4 | 100.00% |
| Together Together | nvidia/nemotron-3-ultra-550b-a55b-20260604 | $0.600 | $3.60 | 512K | — | unknown | 98.98% |
| DeepInfra DeepInfra | nvidia/nemotron-3-ultra-550b-a55b-20260604 | $0.500 | $2.20 | 262K | 16K | fp4 | — |
Benchmarks
Quality index 60.0 / 100
Artificial Analysis
intelligence index
37.8 / 100
coding index
49.3 / 100
agentic index
27.4 / 100
Design Arena
elo models 3d
1212.0
elo models gamedev
1195.0
elo models uicomponent
1179.0
elo models codecategories
1174.0
elo models dataviz
1164.0
elo models website
1140.0
elo models svg
1135.0
elo models asciiart
1105.0
Performance History
53 snapshots
Price History
0 recorded changes
No price changes recorded yet.
Usage (OpenRouter)
Historical usage
Activity Timeline
- No events yet.
Versus
Nearest peers by price, context, and capabilities. One click to compare.
No comparable models found.
First seen 3d ago · Last verified 3m ago