Skip to content
fedi.software

GLM 4.5 in fedi-index

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

fedi-index · General

B 1,396 ±3

Percentile 66 · #100 · provisional

People prefer / Solves tasks

1,408 / —

Human votes vs benchmarks, Elo-eq

Value

+18

$1 per 1M tokens, median of 15 providers · vendor $1

On your hardware

  • 128 GB unified memory #40 · UD-IQ1_M · strong compression · ≈10 tok/s

Score decomposition

B General 1,396 [1,394–1,399] Percentile 66 · #100 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,430 ±2.51 1,430 ±2 26.2% 2026-10-02 —
LMArena text/hard_prompts 1,430 ±3.26 1,422 ±3 26.0% 2026-10-02 —
LMArena text/instruction_following 1,404 ±4.03 1,417 ±4 12.7% 2026-10-02 —
LMArena text/expert 1,432 ±8.91 1,421 ±7 10.9% 2026-10-02 —
LMArena text/longer_query 1,413 ±4.37 1,412 ±4 6.3% 2026-10-02 —
LMArena text/multi_turn 1,415 ±5.04 1,412 ±5 6.2% 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 Coding 1,391 [1,384–1,398] Percentile 63 · #104 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,434 ±4.48 1,412 ±4 90.0% 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

B Math 1,390 [1,376–1,402] Percentile 62 · #97 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,427 ±7.98 1,419 ±8 86.4% 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 1,405 [1,398–1,412] Percentile 66 · #91 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,395 ±5.53 1,423 ±6 61.0% 2026-10-02 —
LMArena text/longer_query 1,413 ±4.37 1,412 ±4 32.4% 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: 繁體中文 1,410 [1,400–1,421] Percentile 65 · #78 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,465 ±7.38 1,425 ±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

B Language: Deutsch 1,389 [1,369–1,409] Percentile 50 · #72 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,406 ±13 1,408 ±11 91.1% 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 1,413 [1,406–1,420] Percentile 68 · #88 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,437 ±3.35 1,427 ±4 93.8% 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: Español 1,427 [1,407–1,447] Percentile 74 · #39 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,454 ±12 1,448 ±11 91.2% 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: Français 1,385 [1,358–1,411] Percentile 38 · #84 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,422 ±16 1,407 ±15 88.9% 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

B Language: 日本語 1,420 [1,405–1,436] Percentile 71 · #38 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,416 ±12 1,439 ±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

B Language: 한국어 1,404 [1,387–1,421] Percentile 55 · #62 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,382 ±12 1,421 ±9 92.0% 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

B Language: Polski 1,411 [1,399–1,422] Percentile 53 · #62 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,422 ±7.58 1,426 ±6 93.1% 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 1,405 [1,393–1,417] Percentile 63 · #87 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,414 ±7.30 1,420 ±7 93.1% 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
GLM 4.5 1,396 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: GLM 4.5

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