Nemotron Nano 8B
8B hybrid Mamba-Transformer reasoning model for PC and edge
About
Nemotron Nano is the 8B entry in NVIDIA's Llama Nemotron family, distilled from Llama 3.1 8B and built as a hybrid Mamba-Transformer for efficiency on PCs and edge devices. It keeps the family's reasoning toggle: NVIDIA reports MATH500 leaping from 36.6% to 95.4% when reasoning mode is switched on. That makes it one of the stronger small reasoning models for local deployment. The trade-off is the usual small-model one — it cannot match the Super or Ultra sizes on hard agentic tasks — and it ships under NVIDIA's Open Model License, not an OSI-approved license.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
NVIDIA-reported: MATH500 95.4 (reasoning on) vs 36.6 (off); MT-Bench ~7.9-8.1
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: Part of NVIDIA Llama Nemotron family (March 18 2025): 8B distilled from Llama 3.1 8B, hybrid Mamba-Transformer.; Benchmarks NVIDIA-reported: MATH500 95.4 reasoning-on vs 36.6 off.; License caveat: NVIDIA Open Model License, not OSI-approved.; Limitation: small-model ceiling on hard agentic tasks..