Skip to content
fedi.software

GPT-4.1 in fedi-index

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

fedi-index · General

C 1,353 ±6

Percentile 48 · #154

People prefer / Solves tasks

1,370 / 1,324

Human votes vs benchmarks, Elo-eq

Value

-56

$3.50 per 1M tokens, median of 23 providers · vendor $3.50

On your hardware

Closed weights: cloud only.

Score decomposition

C General 1,353 [1,346–1,359] Percentile 48 · #154 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 137 ±1.27 1,333 ±10 31.2% 2025-04-14 —
LMArena text/overall 1,383 ±1.89 1,383 ±2 16.9% 2026-10-02 —
LMArena text/hard_prompts 1,384 ±2.49 1,380 ±2 16.8% 2026-10-02 —
LMArena text/instruction_following 1,367 ±2.93 1,382 ±3 8.3% 2026-10-02 —
LMArena text/expert 1,364 ±5.94 1,365 ±5 7.8% 2026-10-02 —
LMArena text/longer_query 1,385 ±3.26 1,387 ±3 4.1% 2026-10-02 —
LMArena text/multi_turn 1,398 ±3.49 1,397 ±3 4.1% 2026-10-02 —
Epoch AI simpleqa_verified outlier decay ×0.61 0.31 ±0.01 1,395 ±5 2.1% 2026-08-31 —
Epoch AI gpqa_diamond decay ×0.26 0.67 ±0.03 1,351 ±13 1.1% 2025-04-14 —

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 1,352 [1,346–1,358] Percentile 46 · #152 · components: 2
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,391 ±3.37 1,373 ±3 78.9% 2026-10-02 —
Epoch AI swe_bench_verified decay ×0.37 0.49 ±0.02 1,322 ±12 12.6% 2026-02-08 —

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 1,340 [1,334–1,345] Percentile 43 · #146 · components: 2
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,369 ±5.23 1,364 ±5 48.5% 2026-10-02 —
Epoch AI frontiermath_t1_3 decay ×0.86 0.06 ±0.01 1,334 ±3 44.7% 2026-08-27 —

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 1,381 [1,376–1,385] Percentile 41 · #67 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,210 ±3.40 1,394 ±3 64.0% 2026-10-02 —
LMArena vision/ocr 1,218 ±4.54 1,392 ±4 15.4% 2026-10-02 —
LMArena vision/diagram 1,215 ±6.62 1,388 ±6 14.4% 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

B Writing 1,378 [1,373–1,383] Percentile 57 · #117 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,364 ±3.92 1,391 ±4 61.9% 2026-10-02 —
LMArena text/longer_query 1,385 ±3.26 1,387 ±3 31.9% 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: 繁體中文 1,350 [1,342–1,358] Percentile 38 · #139 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,382 ±5.41 1,360 ±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

C Language: Deutsch 1,372 [1,359–1,385] Percentile 42 · #84 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,380 ±8.47 1,385 ±7 92.8% 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

B Language: English 1,369 [1,364–1,374] Percentile 54 · #126 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,392 ±2.46 1,380 ±3 94.0% 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: Español 1,363 [1,347–1,378] Percentile 28 · #108 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,376 ±9.28 1,376 ±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

C Language: Français 1,356 [1,338–1,375] Percentile 25 · #101 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,382 ±11 1,371 ±10 91.5% 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

C Language: 日本語 1,355 [1,342–1,368] Percentile 32 · #88 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,319 ±9.56 1,366 ±7 92.9% 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: 한국어 1,376 [1,362–1,389] Percentile 36 · #87 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,339 ±9.65 1,389 ±7 92.8% 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: Polski 1,385 [1,376–1,394] Percentile 35 · #85 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,389 ±5.88 1,398 ±5 93.5% 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

C Language: Русский 1,374 [1,366–1,382] Percentile 48 · #123 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,377 ±4.80 1,386 ±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

History

fedi-index by month
The same in a table
Model 2026-10
GPT-4.1 1,353 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: GPT-4.1

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