Skip to content
fedi.software

GLM 5.3 in fedi-index

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

fedi-index · General

A 1,457 ±6

Percentile 92 · #24

People prefer / Solves tasks

1,460 / 1,453

Human votes vs benchmarks, Elo-eq

Value

+60

$2.15 per 1M tokens, median of 57 providers · vendor $2.15

On your hardware

Does not fit the reference setups.

Score decomposition

A General 1,457 [1,451–1,463] Percentile 92 · #24 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 156 ±1.10 1,482 ±9 33.3% 2026-08-14 —
LMArena text/overall 1,471 ±2.82 1,471 ±3 16.4% 2026-10-02 max
LMArena text/hard_prompts 1,488 ±3.31 1,475 ±3 16.3% 2026-10-02 max
LMArena text/instruction_following 1,477 ±4.21 1,486 ±4 7.9% 2026-10-02 max
LMArena text/expert 1,514 ±7.23 1,488 ±6 7.3% 2026-10-02 max
LMArena text/longer_query 1,483 ±3.80 1,478 ±4 4.0% 2026-10-02 max
LMArena text/multi_turn 1,472 ±6.06 1,464 ±6 3.7% 2026-10-02 max
Epoch AI simpleqa_verified outlier decay ×0.61 0.41 ±0.02 1,426 ±5 2.0% 2026-08-28 max
Epoch AI gpqa_diamond decay ×0.26 0.91 ±0.02 1,463 ±7 1.6% 2026-08-24 max

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

A Coding 1,450 [1,446–1,453] Percentile 94 · #19 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,623 ±4.17 1,485 ±1 46.7% 2026-10-01 max
LMArena text/coding 1,497 ±4.75 1,468 ±4 43.5% 2026-10-02 max

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 Agents 1,442 [1,439–1,445] Percentile 60 · #24 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena agent/overall 0.02 ±0.00 1,467 ±2 90.5% 2026-10-02 max

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

A Math 1,448 [1,438–1,459] Percentile 91 · #25 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,491 ±11 1,480 ±10 45.9% 2026-10-02 max
Epoch AI frontiermath_t1_3 decay ×0.86 0.69 ±0.03 1,470 ±6 33.4% 2026-08-25 max
Epoch AI frontiermath_t4 decay ×0.86 0.29 ±0.07 1,439 ±12 12.2% 2026-08-25 max

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

A Writing 1,466 [1,459–1,472] Percentile 92 · #21 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,457 ±5.21 1,487 ±5 61.0% 2026-10-02 max
LMArena text/longer_query 1,483 ±3.80 1,478 ±4 32.5% 2026-10-02 max

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

A Language: 繁體中文 1,454 [1,441–1,466] Percentile 91 · #21 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,524 ±8.53 1,472 ±7 93.0% 2026-10-02 max

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

A Language: English 1,458 [1,450–1,465] Percentile 94 · #17 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,482 ±3.85 1,475 ±4 93.7% 2026-10-02 max

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: Español 1,419 [1,399–1,440] Percentile 65 · #53 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,446 ±12 1,441 ±12 91.0% 2026-10-02 max

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

A Language: Français 1,447 [1,425–1,469] Percentile 84 · #22 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,494 ±13 1,473 ±12 90.5% 2026-10-02 max

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

A Language: 日本語 1,435 [1,412–1,458] Percentile 81 · #25 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,445 ±17 1,461 ±13 90.1% 2026-10-02 max

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

A Language: Русский 1,442 [1,431–1,453] Percentile 85 · #36 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,458 ±6.58 1,459 ±6 93.3% 2026-10-02 max

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
GLM 5.3 1,457 A

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: GLM 5.3

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