Skip to content
fedi.software

GPT-5 in fedi-index

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

fedi-index · General

B 1,404 ±6

Percentile 70 · #87

People prefer / Solves tasks

1,395 / 1,417

Human votes vs benchmarks, Elo-eq

Value

-4

$3.44 per 1M tokens, median of 27 providers · vendor $3.44

On your hardware

Closed weights: cloud only.

Score decomposition

B General 1,404 [1,397–1,410] Percentile 70 · #87 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 150 1,438 ±11 29.9% 2025-08-07 —
LMArena text/overall 1,406 ±2.26 1,406 ±2 16.8% 2026-10-02 high
LMArena text/hard_prompts 1,415 ±2.93 1,409 ±3 16.7% 2026-10-02 high
LMArena text/instruction_following 1,388 ±3.62 1,402 ±3 8.2% 2026-10-02 high
LMArena text/expert 1,420 ±7.62 1,411 ±6 7.3% 2026-10-02 high
Epoch AI simpleqa_verified decay ×0.61 0.50 ±0.02 1,455 ±5 4.1% 2026-08-27 high
LMArena text/longer_query 1,399 ±3.83 1,400 ±4 4.1% 2026-10-02 —
LMArena text/multi_turn 1,426 ±4.24 1,423 ±4 4.0% 2026-10-02 —
Epoch AI gpqa_diamond decay ×0.26 0.86 ±0.02 1,441 ±10 1.4% 2025-10-29 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 1,406 [1,401–1,410] Percentile 70 · #83 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,417 ±7.92 1,426 ±2 39.6% 2026-10-01 medium
LMArena text/coding 1,436 ±4.04 1,413 ±4 38.3% 2026-10-02 high
Epoch AI swe_bench_verified decay ×0.37 0.74 ±0.02 1,450 ±10 13.6% 2026-02-06 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 1,400 [1,392–1,408] Percentile 71 · #74 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,406 ±7.02 1,399 ±7 50.8% 2026-10-02 —
Epoch AI frontiermath_t1_3 decay ×0.86 0.55 ±0.03 1,441 ±6 29.7% 2026-06-10 high
Epoch AI frontiermath_t4 decay ×0.86 0.22 ±0.07 1,428 ±10 11.8% 2026-06-11 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 1,402 [1,398–1,407] Percentile 46 · #61 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,232 ±3.59 1,413 ±3 63.9% 2026-10-02 —
LMArena vision/ocr 1,245 ±4.61 1,419 ±4 15.4% 2026-10-02 —
LMArena vision/diagram 1,252 ±6.78 1,421 ±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 1,383 [1,377–1,389] Percentile 58 · #114 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,365 ±4.87 1,393 ±5 61.4% 2026-10-02 —
LMArena text/longer_query 1,399 ±3.83 1,400 ±4 32.2% 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: 繁體中文 1,380 [1,371–1,389] Percentile 48 · #117 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,423 ±6.34 1,392 ±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

B Language: Deutsch 1,397 [1,380–1,415] Percentile 56 · #64 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,414 ±11 1,415 ±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

B Language: English 1,385 [1,380–1,391] Percentile 58 · #115 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,409 ±2.99 1,397 ±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 1,382 [1,364–1,399] Percentile 35 · #97 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,399 ±10 1,397 ±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: Français 1,375 [1,351–1,399] Percentile 32 · #92 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,409 ±15 1,395 ±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

B Language: 日本語 1,415 [1,397–1,434] Percentile 66 · #44 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,411 ±14 1,435 ±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: 한국어 1,390 [1,375–1,406] Percentile 47 · #73 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,361 ±11 1,406 ±8 92.4% 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 Language: Polski 1,418 [1,406–1,429] Percentile 63 · #49 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,431 ±7.44 1,434 ±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

B Language: Русский 1,399 [1,388–1,409] Percentile 60 · #94 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,407 ±6.34 1,413 ±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
GPT-5 1,404 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

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