nvidia Open weights Reasoning

Nemotron 3 Super pricing & benchmarks

Released Mar 11, 2026 · served by 3 providers

Input / 1M tokens$0.085
Output / 1M tokens$0.400
Context window1M
Cached input
Intelligence index25.4
Coding index37.7
Agentic index8.7
1M in + 300K out$0.20

What Nemotron 3 Super is

NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. Built on a hybrid Mamba-Transformer...

Providers and prices for Nemotron 3 Super

The same model costs different amounts depending on who serves it. Prices are per 1M tokens, cheapest first. Uptime and throughput are OpenRouter's rolling measurements, not vendor claims.

ProviderInputOutputContext QuantUptime 24hThroughput
DeepInfracheapest $0.085 $0.400 262K bf16 91.0%
DigitalOcean $0.210 $0.455 1M 96.4%
Nebius $0.300 $0.900 262K fp4 99.7%

Cheaper models in the same class

Models scoring within 4 points of Nemotron 3 Super on the Intelligence Index, but with a lower output price.

FAQ

How much does the Nemotron 3 Super API cost?

$0.085 per 1M input tokens and $0.400 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $0.20.

Is Nemotron 3 Super free?

No. It is a paid model starting at $0.085 per 1M input tokens, though some providers offer trial credits.

Which provider is cheapest for Nemotron 3 Super?

DeepInfra at $0.085 input / $0.400 output per 1M tokens (bf16 quantization).

Can I self-host Nemotron 3 Super?

Yes — weights are published as nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8 on Hugging Face, so you can run it on your own hardware.