Skip to content
fedi.software

gemma-3n-e4b-it in fedi-index

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

fedi-index · General

C 1,279 ±3

Percentile 31 · #203 · provisional

People prefer / Solves tasks

1,284 / —

Human votes vs benchmarks, Elo-eq

Value

Free

On your hardware

Does not fit the reference setups.

Score decomposition

C General 1,279 [1,276–1,282] Percentile 31 · #203 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,305 ±2.64 1,306 ±3 26.4% 2026-10-02 —
LMArena text/hard_prompts 1,284 ±3.62 1,290 ±3 26.0% 2026-10-02 —
LMArena text/instruction_following 1,255 ±4.39 1,276 ±4 12.6% 2026-10-02 —
LMArena text/expert 1,245 ±9.26 1,268 ±8 10.8% 2026-10-02 —
LMArena text/multi_turn 1,276 ±5.35 1,287 ±5 6.2% 2026-10-02 —
LMArena text/longer_query 1,277 ±5.59 1,286 ±5 6.1% 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 Coding 1,258 [1,250–1,266] Percentile 28 · #201 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,269 ±5.23 1,264 ±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 1,244 [1,231–1,256] Percentile 29 · #181 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,251 ±7.65 1,250 ±7 86.6% 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 1,297 [1,289–1,304] Percentile 36 · #172 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,287 ±5.68 1,313 ±6 61.7% 2026-10-02 —
LMArena text/longer_query 1,277 ±5.59 1,286 ±5 31.6% 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 Language: 繁體中文 1,296 [1,284–1,307] Percentile 28 · #162 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,310 ±7.98 1,303 ±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: Deutsch 1,316 [1,299–1,333] Percentile 27 · #105 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,312 ±11 1,326 ±10 91.9% 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: English 1,298 [1,292–1,305] Percentile 35 · #177 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,321 ±3.39 1,305 ±4 93.8% 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

D Language: Español 1,300 [1,276–1,324] Percentile 10 · #134 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,305 ±15 1,311 ±14 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

D Language: Français 1,309 [1,282–1,336] Percentile 14 · #116 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,330 ±17 1,323 ±16 88.6% 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

D Language: 日本語 1,322 [1,307–1,338] Percentile 24 · #98 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,273 ±12 1,332 ±9 92.3% 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

D Language: 한국어 1,317 [1,298–1,335] Percentile 19 · #110 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,258 ±14 1,327 ±10 91.6% 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

D Language: Polski 1,303 [1,292–1,314] Percentile 4 · #126 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,285 ±7.26 1,310 ±6 93.2% 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 1,299 [1,287–1,311] Percentile 29 · #167 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,288 ±7.20 1,306 ±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

History

fedi-index by month
The same in a table
Model 2026-10
gemma-3n-e4b-it 1,279 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: gemma-3n-e4b-it

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