Skip to content
fedi.software

Qwen3 Max in fedi-index

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

fedi-index · General

B 1,400 ±6

Percentile 68 · #93

People prefer / Solves tasks

1,423 / 1,365

Human votes vs benchmarks, Elo-eq

Value

+2

$2.30 per 1M tokens, median of 20 providers · vendor $2.40

On your hardware

Closed weights: cloud only.

Score decomposition

B General 1,400 [1,394–1,406] Percentile 68 · #93 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 142 ±1.18 1,378 ±9 32.6% 2025-09-24 —
LMArena text/overall 1,439 ±2.31 1,439 ±2 16.8% 2026-10-02 —
LMArena text/hard_prompts 1,448 ±2.97 1,438 ±3 16.6% 2026-10-02 —
LMArena text/instruction_following 1,420 ±3.74 1,432 ±4 8.1% 2026-10-02 —
LMArena text/expert 1,457 ±8.31 1,441 ±7 7.1% 2026-10-02 —
LMArena text/longer_query 1,439 ±4.04 1,436 ±4 4.0% 2026-10-02 —
LMArena text/multi_turn 1,446 ±4.62 1,441 ±4 4.0% 2026-10-02 —
Epoch AI simpleqa_verified outlier decay ×0.61 0.49 ±0.02 1,450 ±5 2.1% 2026-08-27 —
Epoch AI gpqa_diamond decay ×0.26 0.73 ±0.03 1,378 ±13 1.2% 2025-10-06 —

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 1,408 [1,402–1,415] Percentile 73 · #77 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,456 ±4.26 1,431 ±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 1,386 [1,378–1,394] Percentile 60 · #102 · components: 2
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,447 ±7.63 1,438 ±7 47.3% 2026-10-02 —
Epoch AI frontiermath_t1_3 decay ×0.86 0.19 ±0.02 1,362 ±5 45.4% 2026-08-30 —

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 1,417 [1,411–1,424] Percentile 73 · #73 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,401 ±5.08 1,430 ±5 61.3% 2026-10-02 —
LMArena text/longer_query 1,439 ±4.04 1,436 ±4 32.3% 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: 繁體中文 1,422 [1,412–1,432] Percentile 74 · #58 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,480 ±6.72 1,437 ±5 93.4% 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

A Language: Deutsch 1,434 [1,415–1,454] Percentile 82 · #26 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,463 ±13 1,457 ±11 91.3% 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: English 1,418 [1,412–1,425] Percentile 72 · #76 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,442 ±3.11 1,433 ±3 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

A Language: Español 1,436 [1,418–1,453] Percentile 80 · #30 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,462 ±11 1,456 ±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: Français 1,409 [1,386–1,433] Percentile 52 · #66 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,450 ±15 1,433 ±14 89.9% 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 1,407 [1,387–1,426] Percentile 59 · #53 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,399 ±15 1,426 ±11 91.3% 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 1,415 [1,399–1,432] Percentile 65 · #48 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,397 ±12 1,433 ±9 92.1% 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: Polski 1,424 [1,412–1,437] Percentile 69 · #41 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,440 ±8.09 1,441 ±7 93.0% 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: Русский 1,416 [1,405–1,427] Percentile 70 · #72 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,428 ±6.87 1,432 ±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

History

fedi-index by month
The same in a table
Model 2026-10
Qwen3 Max 1,400 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: Qwen3 Max

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