Skip to content
fedi.software

GLM 4.6 in fedi-index

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

fedi-index · General

B 1,404 ±2

Percentile 70 · #88 · provisional

People prefer / Solves tasks

1,416 / —

Human votes vs benchmarks, Elo-eq

Value

+26

$0.9813 per 1M tokens, median of 20 providers · vendor $1

On your hardware

  • 128 GB unified memory #38 · UD-IQ1_M · strong compression · ≈10 tok/s

Score decomposition

B General 1,404 [1,401–1,406] Percentile 70 · #88 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,440 ±1.99 1,440 ±2 25.9% 2026-10-02 —
LMArena text/hard_prompts 1,440 ±2.53 1,431 ±2 25.8% 2026-10-02 —
LMArena text/instruction_following 1,410 ±3.22 1,422 ±3 12.6% 2026-10-02 —
LMArena text/expert 1,431 ±6.61 1,420 ±5 11.7% 2026-10-02 —
LMArena text/longer_query 1,422 ±3.35 1,421 ±3 6.3% 2026-10-02 —
LMArena text/multi_turn 1,427 ±4.16 1,423 ±4 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 1,394 [1,390–1,397] Percentile 66 · #96 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,339 ±5.49 1,404 ±2 46.0% 2026-10-01 —
LMArena text/coding 1,449 ±3.65 1,425 ±3 44.4% 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 Math 1,395 [1,384–1,406] Percentile 67 · #84 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,431 ±6.41 1,423 ±6 87.3% 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 Writing 1,419 [1,413–1,425] Percentile 75 · #68 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,411 ±4.33 1,440 ±4 61.6% 2026-10-02 —
LMArena text/longer_query 1,422 ±3.35 1,421 ±3 32.1% 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

A Language: 繁體中文 1,437 [1,428–1,445] Percentile 82 · #42 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,499 ±5.84 1,453 ±5 93.6% 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 1,423 [1,405–1,440] Percentile 73 · #40 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,447 ±11 1,442 ±10 91.8% 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 1,425 [1,420–1,430] Percentile 77 · #64 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,448 ±2.74 1,439 ±3 93.9% 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: Español 1,414 [1,398–1,430] Percentile 59 · #62 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,436 ±9.71 1,432 ±9 92.2% 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: Français 1,421 [1,402–1,440] Percentile 66 · #47 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,459 ±12 1,441 ±11 91.4% 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: 日本語 1,402 [1,382–1,423] Percentile 54 · #59 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,394 ±15 1,422 ±12 90.9% 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: 한국어 1,416 [1,399–1,434] Percentile 67 · #46 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,400 ±13 1,435 ±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

B Language: Polski 1,430 [1,417–1,442] Percentile 75 · #33 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,447 ±8.28 1,447 ±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

B Language: Русский 1,409 [1,401–1,417] Percentile 66 · #79 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,418 ±4.90 1,423 ±4 93.6% 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 4.6 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: GLM 4.6

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