Skip to content
fedi.software

GPT-5.4 Nano in fedi-index

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

fedi-index · General

B 1,373 ±6

Percentile 54 · #135

People prefer / Solves tasks

1,367 / 1,383

Human votes vs benchmarks, Elo-eq

Value

+14

$0.4625 per 1M tokens, median of 28 providers · vendor $0.4625

On your hardware

Closed weights: cloud only.

Score decomposition

B General 1,373 [1,368–1,379] Percentile 54 · #135 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 146 ±1.06 1,405 ±8 33.1% 2026-03-17 —
LMArena text/overall 1,372 ±1.92 1,372 ±2 16.3% 2026-10-02 high
LMArena text/hard_prompts 1,381 ±2.31 1,377 ±2 16.2% 2026-10-02 high
LMArena text/instruction_following 1,362 ±2.81 1,377 ±3 8.0% 2026-10-02 high
LMArena text/expert 1,396 ±4.29 1,391 ±4 7.8% 2026-10-02 high
LMArena text/longer_query 1,366 ±2.72 1,368 ±2 4.0% 2026-10-02 high
LMArena text/multi_turn 1,382 ±3.45 1,383 ±3 4.0% 2026-10-02 high
Epoch AI simpleqa_verified outlier decay ×0.61 0.12 ±0.01 1,334 ±3 2.1% 2026-08-27 high
Epoch AI gpqa_diamond decay ×0.26 0.78 ±0.02 1,405 ±11 1.2% 2026-04-14 high

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

B Coding 1,368 [1,363–1,372] Percentile 52 · #134 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,405 ±3.01 1,385 ±3 90.3% 2026-10-02 high

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

B Math 1,392 [1,385–1,399] Percentile 64 · #93 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,406 ±5.54 1,399 ±5 51.6% 2026-10-02 high
Epoch AI frontiermath_t1_3 decay ×0.86 0.45 ±0.03 1,418 ±6 28.3% 2026-06-12 high
Epoch AI frontiermath_t4 decay ×0.86 0.12 ±0.05 1,412 ±8 12.8% 2026-06-12 high

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 Vision 1,370 [1,366–1,374] Percentile 37 · #72 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,196 ±3.23 1,381 ±3 63.0% 2026-10-02 high
LMArena vision/ocr 1,204 ±3.24 1,379 ±3 15.7% 2026-10-02 high
LMArena vision/diagram 1,211 ±4.55 1,384 ±4 15.2% 2026-10-02 high

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 1,341 [1,336–1,346] Percentile 47 · #143 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,314 ±3.67 1,340 ±4 61.9% 2026-10-02 high
LMArena text/longer_query 1,366 ±2.72 1,368 ±2 32.0% 2026-10-02 high

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 1,356 [1,349–1,364] Percentile 40 · #133 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,391 ±5.33 1,367 ±4 93.7% 2026-10-02 high

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 1,361 [1,345–1,378] Percentile 37 · #91 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,368 ±11 1,375 ±9 92.1% 2026-10-02 high

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 1,358 [1,353–1,363] Percentile 49 · #139 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,381 ±2.54 1,368 ±3 93.9% 2026-10-02 high

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: Español 1,359 [1,346–1,372] Percentile 25 · #112 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,370 ±7.68 1,370 ±7 92.9% 2026-10-02 high

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 1,370 [1,358–1,383] Percentile 30 · #95 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,396 ±7.63 1,383 ±7 92.9% 2026-10-02 high

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: 日本語 1,368 [1,350–1,386] Percentile 37 · #81 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,341 ±13 1,383 ±10 91.8% 2026-10-02 high

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: 한국어 1,363 [1,349–1,377] Percentile 33 · #91 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,322 ±10 1,376 ±8 92.7% 2026-10-02 high

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 1,365 [1,351–1,379] Percentile 24 · #100 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,365 ±9.31 1,378 ±8 92.6% 2026-10-02 high

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 1,362 [1,356–1,369] Percentile 44 · #132 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,362 ±4.07 1,373 ±4 93.8% 2026-10-02 high

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
GPT-5.4 Nano 1,373 B

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: GPT-5.4 Nano

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