Skip to content
fedi.software

ernie-5.1 in fedi-index

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

fedi-index · General

A 1,437 ±2

Percentile 85 · #45 · provisional

People prefer / Solves tasks

1,451 / —

Human votes vs benchmarks, Elo-eq

Value

+53

$1.31 per 1M tokens, median of 1 providers

On your hardware

Closed weights: cloud only.

Score decomposition

A General 1,437 [1,435–1,439] Percentile 85 · #45 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,468 ±2.31 1,468 ±2 25.7% 2026-10-02 —
LMArena text/hard_prompts 1,481 ±2.70 1,468 ±2 25.6% 2026-10-02 —
LMArena text/instruction_following 1,460 ±3.34 1,470 ±3 12.6% 2026-10-02 —
LMArena text/expert 1,492 ±5.24 1,469 ±4 12.2% 2026-10-02 —
LMArena text/longer_query 1,462 ±3.17 1,458 ±3 6.3% 2026-10-02 —
LMArena text/multi_turn 1,471 ±4.16 1,463 ±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

A Coding 1,435 [1,429–1,440] Percentile 87 · #36 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,488 ±3.52 1,460 ±3 90.2% 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

A Math 1,436 [1,425–1,447] Percentile 88 · #32 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,481 ±6.82 1,471 ±6 87.1% 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 Search 1,452 [1,443–1,462] Percentile 84 · #6 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena search/overall 1,227 ±4.95 1,470 ±5 93.5% 2026-08-24 —

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 1,450 [1,444–1,455] Percentile 86 · #39 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,441 ±4.25 1,470 ±4 61.6% 2026-10-02 —
LMArena text/longer_query 1,462 ±3.17 1,458 ±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 1,442 [1,432–1,452] Percentile 84 · #36 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,507 ±6.86 1,459 ±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 1,441 [1,424–1,459] Percentile 87 · #20 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,470 ±11 1,462 ±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

A Language: English 1,454 [1,448–1,460] Percentile 92 · #22 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,478 ±2.94 1,470 ±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

A Language: Español 1,446 [1,431–1,462] Percentile 89 · #17 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,473 ±9.10 1,466 ±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

A Language: Français 1,448 [1,433–1,462] Percentile 86 · #20 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,488 ±8.67 1,467 ±8 92.6% 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: 日本語 1,422 [1,404–1,441] Percentile 74 · #34 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,421 ±14 1,443 ±10 91.6% 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

A Language: 한국어 1,436 [1,419–1,453] Percentile 82 · #26 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,428 ±12 1,456 ±10 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

A Language: Polski 1,444 [1,428–1,461] Percentile 87 · #18 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,468 ±10 1,465 ±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

A Language: Русский 1,444 [1,436–1,452] Percentile 87 · #32 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,459 ±4.84 1,460 ±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
ernie-5.1 1,437 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: ernie-5.1

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