Skip to content
fedi.software

Nemotron 3 Nano 30B A3B in fedi-index

The score in every slice, what it is made of, value, local run and history.

fedi-index · General

C 1,324 ±3

Percentile 40 · #177 · provisional

People prefer / Solves tasks

1,332 / —

Human votes vs benchmarks, Elo-eq

Value

+6

$0.0875 per 1M tokens, median of 13 providers

On your hardware

  • GPU 12 GB #12 · Q8_0 · ≈16 tok/s
  • GPU 24 GB #15 · Q4_K_M · ≈170 tok/s
  • Mac 64 GB #14 · Q8_0 · ≈44 tok/s
  • 128 GB unified memory #22 · Q8_0 · ≈33 tok/s
  • CPU, 32 GB RAM #7 · Q4_K_M · ≈7 tok/s

Score decomposition

C General 1,324 [1,321–1,327] Percentile 40 · #177 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,348 ±2.76 1,348 ±3 26.4% 2026-10-02 —
LMArena text/hard_prompts 1,348 ±3.58 1,348 ±3 26.1% 2026-10-02 —
LMArena text/instruction_following 1,304 ±4.64 1,322 ±4 12.6% 2026-10-02 —
LMArena text/expert 1,354 ±9.44 1,357 ±8 10.8% 2026-10-02 —
LMArena text/longer_query 1,300 ±4.89 1,307 ±5 6.2% 2026-10-02 —
LMArena text/multi_turn 1,319 ±6.02 1,326 ±6 6.0% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Coding 1,346 [1,338–1,354] Percentile 42 · #162 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,378 ±5.20 1,362 ±5 89.8% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Math 1,341 [1,326–1,356] Percentile 44 · #144 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,369 ±9.24 1,364 ±9 85.6% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Writing 1,293 [1,284–1,301] Percentile 35 · #174 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,269 ±6.36 1,295 ±6 60.4% 2026-10-02 —
LMArena text/longer_query 1,300 ±4.89 1,307 ±5 32.8% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Language: 繁體中文 1,349 [1,337–1,361] Percentile 37 · #140 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,382 ±8.21 1,360 ±6 93.1% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Language: English 1,354 [1,346–1,361] Percentile 47 · #146 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,377 ±3.97 1,364 ±4 93.7% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Language: Español 1,358 [1,332–1,384] Percentile 24 · #114 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,377 ±16 1,377 ±15 88.9% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Language: Français 1,364 [1,339–1,389] Percentile 28 · #98 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,396 ±15 1,383 ±14 89.5% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Language: Polski 1,314 [1,294–1,335] Percentile 7 · #122 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,303 ±13 1,325 ±11 91.0% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

C Language: Русский 1,312 [1,302–1,323] Percentile 32 · #160 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,303 ±6.39 1,320 ±6 93.3% 2026-10-02 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

History

fedi-index by month
The same in a table
Model 2026-10
Nemotron 3 Nano 30B A3B 1,324 C

Badge

Show the model's fedi-index on your site or in a README. The badge updates itself; it links to this page.

fedi-index: Nemotron 3 Nano 30B A3B

Data CC BY 4.0: keep the link to fedi.software.