Nemotron 3 Ultra pricing & benchmarks
Nemotron 3 Ultra is available from NVIDIA at $0.600 per million input tokens and $3.60 per million output tokens ($1.35 blended at 3:1). It accepts up to 512,288 tokens of context. 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...
Pricing
Per million tokens · OpenRouter
Input$0.600
Output$3.60
Blended (3:1)$1.35
Specification
nvidia/nemotron-3-ultra-550b-a55b
Context window512K
Max output—
ReleasedJun 2026
LicenceOpen weights
Modalitiesin:text, out:text
No independently-run benchmark scores are published for Nemotron 3 Ultra yet. We show nothing rather than reprinting an unverified figure.
Where to run Nemotron 3 Ultra
4 hosts · 1.5× between cheapest and dearest. Blended at 3:1. ⚠ marks an option serving less context than the best available, or re-pricing above a prompt-length threshold. Bracketed labels are the provider's own service tier or region.
| Host | Blended $/1M | In / Out | Context | Weights | Uptime 24h |
|---|---|---|---|---|---|
| DeepInfra ⚠ 262K context, not 512K |
$0.925 | $0.500 / $2.20 | 262K | fp4 | 85.3% |
| BaseTen ⚠ 203K context, not 512K |
$1.05 | $0.600 / $2.40 | 203K | fp4 | 100.0% |
| Venice ⚠ 256K context, not 512K |
$1.25 | $0.625 / $3.13 | 256K | fp8 | 86.5% |
| Together | $1.35 | $0.600 / $3.60 | 512K | — | 98.1% |
Host prices and uptime from OpenRouter, synced 2026-08-08. Lower-precision weights (fp4, fp8) can cost less and score worse than the same model at bf16 — the score above is not host-specific.
Compare Nemotron 3 Ultra
FAQ
How much does Nemotron 3 Ultra cost?
Nemotron 3 Ultra costs $0.600 per million input tokens and $3.60 per million output tokens on OpenRouter. At a 3:1 input:output mix that blends to $1.35 per million tokens.
What is Nemotron 3 Ultra's context window?
Nemotron 3 Ultra accepts up to 512,288 tokens of context.