NVIDIA
Nemotron 3 Ultra
NVIDIA released Nemotron 3 Ultra on June 4, 2026 with 550B total and 55B active parameters under the OpenMDW-1.1 license. It is text-only, with context up to 1M tokens and an enable_thinking toggle for reasoning through its chat template. NVIDIA does not publish a per-token list price for Nemotron 3 Ultra; hosted pricing varies by provider.
1M tokens
Text only
Pro
- Provider
- NVIDIA
- Context
- 1MCatalog · As of Oct 21, 2026Under 1d old
- yno subscription tier
- Pro
- Released
- Jun 4, 2026
- Speed
- Medium
- Reasoning
- Expert
- Modality
- Text only
Benchmarks
Quality
- AA Intelligence Index
- AA Coding Index
- GPQA Diamond
- 86.7%Artificial Analysis · Date unknownAge unknown
Speed and latency
- Output speed
- 159.6tokens/sConfiguration: Prompt length: 1,000 · Parallel queries: 1Artificial Analysis · Date unknownAge unknown
- Time to first token
- 0.81sConfiguration: Prompt length: 1,000 · Parallel queries: 1Artificial Analysis · Date unknownAge unknown
- Time to first answer token
- 15.07sConfiguration: Prompt length: 1,000 · Parallel queries: 1Artificial Analysis · Date unknownAge unknown
API pricing
- API input cost
- API output cost
Benchmark variant: Nemotron 3 Ultra 550B A55B (Reasoning)
AA data retrieved Oct 21, 2026 · Artificial Analysis
Indexes use points; evaluations use accuracy percentages. API measurements do not measure yno application performance. Methodology
Capabilities
- Open Source
- Reasoning
- Long Context
- Agentic
Use cases
- Long-running agents
- Text analysis
- Self-hosted AI
- Reasoning workloads
Strengths
- OpenMDW-1.1 weights
- 550B total / 55B active parameters
- Context up to 1M tokens
- Configurable enable_thinking toggle
Best for
Teams deploying text-only reasoning agents with configurable thinking and published weights
Use Nemotron 3 Ultra in yno.ai
No credit card required
Related models
Frequently asked
- What is Nemotron 3 Ultra?
- NVIDIA released Nemotron 3 Ultra on June 4, 2026 with 550B total and 55B active parameters under the OpenMDW-1.1 license. It is text-only, with context up to 1M tokens and an enable_thinking toggle for reasoning through its chat template. NVIDIA does not publish a per-token list price for Nemotron 3 Ultra; hosted pricing varies by provider.
- How much does Nemotron 3 Ultra cost in yno.ai?
- Nemotron 3 Ultra is available on the Pro tier of yno.ai.
- What can Nemotron 3 Ultra do?
- Nemotron 3 Ultra is best for Teams deploying text-only reasoning agents with configurable thinking and published weights. Its main capabilities include Open Source, Reasoning, Long Context, Agentic.
- How does Nemotron 3 Ultra compare to other models?
- Nemotron 3 Ultra excels at OpenMDW-1.1 weights and is recommended for Long-running agents, Text analysis, Self-hosted AI, Reasoning workloads. See related models below.
- How do I use Nemotron 3 Ultra in yno.ai?
- Sign up for yno.ai, select Nemotron 3 Ultra from the model picker, and start chatting.