Back to models
deprecated

NVIDIA: Nemotron 3 Super

nvidia/nemotron-3-super-120b-a12b

Prices

Input

$0.20 per 1M tokens

low
Output

$0.80 per 1M tokens

low

Also available on

The same model, sold by someone else. The price above is the maker's own.

14 sales channels for this model.
ChannelWhatPrice
OpenRouter · nvidiaInput$0.075 per 1M tokens
Input (cached read)$0.15 per 1M tokens
Output$0.40 per 1M tokens
OpenRouterInput$0.075 per 1M tokens
Input (cached read)$0.15 per 1M tokens
Output$0.40 per 1M tokens
Other · deepinfraInput$0.075 per 1M tokens
Output$0.40 per 1M tokens
Other · edenaiInput (cached read)$0.30 per 1M tokens
Input (host: deepinfra)$0.075 per 1M tokens
Input (host: flexai)$0.085 per 1M tokens
Input (host: nebius)$0.30 per 1M tokens
Output (host: deepinfra)$0.40 per 1M tokens
Output (host: flexai)$0.40 per 1M tokens
Output (host: nebius)$0.90 per 1M tokens
Other · kenariInput$0.0971 per 1M tokens
Input (cached read)$0.009714 per 1M tokens
Output$0.48 per 1M tokens
Other · kiloInput$0.06 per 1M tokens
Output$0.32 per 1M tokens
Other · nano-gptInput$0.05 per 1M tokens
Input (cached read)$0.025 per 1M tokens
Output$0.25 per 1M tokens
Other · nebiusInput$0.30 per 1M tokens
Input (cached read)$0.06 per 1M tokens
Output$0.90 per 1M tokens
Other · openrouterInput$0.08 per 1M tokens
Output$0.45 per 1M tokens
Other · perplexity-agentInput$0.25 per 1M tokens
Output$2.50 per 1M tokens
Other · requestyInput$0.10 per 1M tokens
Output$0.50 per 1M tokens
Other · stepping-stoneInput$2.38 per 1M tokens
Input (cached read)$0.357 per 1M tokens
Output$5.95 per 1M tokens
Other · tensorxInput$0.30 per 1M tokens
Input (cache write)$0.375 per 1M tokens
Input (cached read)$0.075 per 1M tokens
Output$0.90 per 1M tokens
Other · vercelInput$0.15 per 1M tokens
Output$0.65 per 1M tokens

Specifications

Provider
nvidia
Context window
262,144
Max output tokens
262,144
Open weights
yes
Released
03/11/2026
Retires
no date announced
First seen by this tracker

Price history

No price of this model has changed since we started tracking it.

Input
0.2 per 1M tokens, unchanged since 09/14/2026
Output
0.8 per 1M tokens, unchanged since 09/14/2026

Last updated 11/17/2026, 01:12