modelbenchmark.io

nvidia-nemotron-3-nano-30b-a3b

NVIDIA · nvidia-nemotron-3-nano-30b-a3b

Compare

Nemotron-Nano-3-30B-A3B is a compact Mixture-of-Experts model optimized for efficient reasoning, chat, and coding, with strong multilingual support and long-context RAG and agent workflows.

Specification

most-agreed values

Context
256K
Max output
256K
Released
2025-12-15
Knowledge cutoff
2025-06-25
Retires
—
Open weights
yes
Input
text
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.06
Output
$0.24
Cache read
$0.03

Available from 4 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
crusoe$0.05$0.20$0.03—262K262K—
cortecs$0.06$0.24——256K256K—
nebius$0.06$0.24——262K262K—
venice$0.075$0.30——128K16K—

Price history

append-only observations · a listing writes a row only when its price moves

DateHostField$/M
2026-09-02nebiusInput$0.06
2026-09-06nebiusInput$0.06
2026-09-02nebiusOutput$0.24
2026-09-06nebiusOutput$0.24

More from NVIDIA

most-hosted first