Home › Providers › nvidia › NVIDIA: Nemotron 3 Ultra
NVIDIA: Nemotron 3 Ultra API Pricing
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...
How much does NVIDIA: Nemotron 3 Ultra cost for a real workload?
At 10M input + 2M output tokens per month, NVIDIA: Nemotron 3 Ultra costs about $10.8 ($6 input + $4.8 output). Try your own volumes in the cost calculator.
What are cheaper alternatives to NVIDIA: Nemotron 3 Ultra?
- Z.ai: GLM 5.3 — $0.7 in / $2.2 out per 1M
- MoonshotAI: Kimi K2 0711 — $0.57 in / $2.3 out per 1M
- Z.ai: GLM 4.5 — $0.6 in / $2.2 out per 1M
- Moondream: Moondream 3 Preview — $0.3 in / $2.5 out per 1M
- Moondream: Moondream 3.1 — $0.3 in / $2.5 out per 1M
- Google: Gemini 3.5 Flash Lite — $0.3 in / $2.5 out per 1M
Where to run NVIDIA: Nemotron 3 Ultra
AI/ML API: one API key for 400+ models, including most models tracked here · Novita AI: pay-as-you-go APIs for open models (DeepSeek, Qwen, Llama and more) plus GPU cloud. Prices there follow each platform's own catalog and can differ from the list prices tracked here.
Run it on AI/ML API →Run it on Novita AI →
Disclosure: links to these platforms may be affiliate links — if you sign up or buy through them, LLMPrices may earn a commission at no extra cost to you.
Compare NVIDIA: Nemotron 3 Ultra with other models
NVIDIA: Nemotron 3 Ultra vs NVIDIA: Nemotron 3.5 Super VL (free)NVIDIA: Nemotron 3 Ultra vs NVIDIA: Nemotron 3.5 Lightning (free)NVIDIA: Nemotron 3 Ultra vs NVIDIA: Nemotron 3.5 Content Safety (free)NVIDIA: Nemotron 3 Ultra vs NVIDIA: Nemotron 3 Ultra (free)NVIDIA: Nemotron 3 Ultra vs NVIDIA: Nemotron 3 Nano Omni (free)NVIDIA: Nemotron 3 Ultra vs NVIDIA: Nemotron 3 Super (free)
Frequently asked questions
How much does NVIDIA: Nemotron 3 Ultra cost?
NVIDIA: Nemotron 3 Ultra costs $0.6 per 1M input tokens and $2.4 per 1M output tokens (list price, snapshot October 22, 2026). At 10M input + 2M output tokens per month that is about $10.8.
What is the context window of NVIDIA: Nemotron 3 Ultra?
NVIDIA: Nemotron 3 Ultra has a context window of 262,144 tokens and can return up to 182,520 output tokens per response.
What are cheaper alternatives to NVIDIA: Nemotron 3 Ultra?
Cheaper models include Z.ai: GLM 5.3 ($0.7 in / $2.2 out), MoonshotAI: Kimi K2 0711 ($0.57 in / $2.3 out), Z.ai: GLM 4.5 ($0.6 in / $2.2 out). See the full list on this page.
Is NVIDIA: Nemotron 3 Ultra free to use?
No. NVIDIA: Nemotron 3 Ultra is a paid model at $0.6 input / $2.4 output per 1M tokens. 21 free models are listed on the free models page.
Provider: nvidia · Model ID: nvidia/nemotron-3-ultra-550b-a55b · Modality: text->text · Price snapshot October 22, 2026, refreshed daily. How we collect prices