Skip to content
fedi.software

GPT-5 Mini in fedi-index

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

fedi-index · General

B 1,371 ±5

Percentile 53 · #137

People prefer / Solves tasks

1,362 / 1,382

Human votes vs benchmarks, Elo-eq

Value

+2

$0.6875 per 1M tokens, median of 29 providers · vendor $0.6875

On your hardware

Closed weights: cloud only.

Score decomposition

B General 1,371 [1,366–1,376] Percentile 53 · #137 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 146 ±0.91 1,402 ±7 34.5% 2025-08-07 —
LMArena text/overall 1,373 ±2.33 1,373 ±2 15.8% 2026-10-02 high
LMArena text/hard_prompts 1,380 ±3.05 1,377 ±3 15.6% 2026-10-02 high
LMArena text/instruction_following 1,357 ±3.82 1,372 ±4 7.6% 2026-10-02 high
LMArena text/expert 1,377 ±8.89 1,376 ±7 6.6% 2026-10-02 high
Epoch AI simpleqa_verified decay ×0.61 0.22 ±0.01 1,365 ±4 4.0% 2026-08-10 high
LMArena text/longer_query 1,355 ±4.12 1,358 ±4 3.8% 2026-10-02 high
LMArena text/multi_turn 1,363 ±4.69 1,365 ±4 3.7% 2026-10-02 high
Epoch AI gpqa_diamond decay ×0.26 0.75 ±0.02 1,389 ±11 1.2% 2025-10-30 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 1,373 [1,366–1,380] Percentile 54 · #129 · components: 2
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,406 ±4.24 1,386 ±4 78.1% 2026-10-02 high
Epoch AI swe_bench_verified decay ×0.37 0.65 ±0.02 1,405 ±11 13.2% 2026-02-01 medium

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 1,379 [1,370–1,387] Percentile 56 · #112 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,377 ±7.90 1,371 ±8 48.9% 2026-10-02 high
Epoch AI frontiermath_t1_3 decay ×0.86 0.47 ±0.03 1,422 ±6 29.8% 2026-06-12 high
Epoch AI frontiermath_t4 decay ×0.86 0.12 ±0.05 1,412 ±8 13.6% 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 1,371 [1,366–1,376] Percentile 38 · #71 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,202 ±3.98 1,386 ±4 64.2% 2026-10-02 high
LMArena vision/ocr 1,199 ±5.25 1,374 ±5 15.2% 2026-10-02 high
LMArena vision/diagram 1,203 ±7.53 1,377 ±7 14.2% 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 1,345 [1,338–1,352] Percentile 48 · #141 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,325 ±5.24 1,352 ±5 61.2% 2026-10-02 high
LMArena text/longer_query 1,355 ±4.12 1,358 ±4 32.4% 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 1,353 [1,343–1,363] Percentile 40 · #135 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,387 ±7.05 1,364 ±6 93.4% 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 1,358 [1,339–1,377] Percentile 36 · #93 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,365 ±12 1,372 ±11 91.4% 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 1,355 [1,349–1,362] Percentile 48 · #143 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,379 ±3.15 1,365 ±3 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

D Language: Español 1,345 [1,327–1,362] Percentile 19 · #121 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,355 ±11 1,357 ±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

D Language: Français 1,355 [1,331–1,380] Percentile 24 · #103 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,385 ±15 1,373 ±14 89.5% 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 1,367 [1,348–1,387] Percentile 36 · #82 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,341 ±15 1,383 ±11 91.2% 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 Language: 한국어 1,351 [1,334–1,368] Percentile 29 · #97 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,307 ±12 1,364 ±10 92.0% 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: Polski 1,376 [1,364–1,388] Percentile 31 · #91 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,379 ±7.89 1,389 ±7 93.0% 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: Русский 1,361 [1,350–1,372] Percentile 42 · #135 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,361 ±6.69 1,372 ±6 93.2% 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

History

fedi-index by month
The same in a table
Model 2026-10
GPT-5 Mini 1,371 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 Mini

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