Skip to content
fedi.software

Claude Fable 5 in fedi-index

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

fedi-index · General

S 1,484 ±7

Percentile 99 · #5

People prefer / Solves tasks

1,479 / 1,492

Human votes vs benchmarks, Elo-eq

Value

+33

$20 per 1M tokens, median of 32 providers · vendor $20

On your hardware

Closed weights: cloud only.

Score decomposition

S General 1,484 [1,476–1,491] Percentile 99 · #5 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 162 ±1.76 1,533 ±14 25.7% 2026-06-09 —
LMArena text/overall 1,492 ±2.24 1,492 ±2 17.8% 2026-10-02 high
LMArena text/hard_prompts 1,508 ±2.58 1,493 ±2 17.7% 2026-10-02 high
LMArena text/instruction_following 1,503 ±3.14 1,511 ±3 8.7% 2026-10-02 high
LMArena text/expert 1,534 ±4.91 1,504 ±4 8.5% 2026-10-02 high
Epoch AI simpleqa_verified decay ×0.61 0.71 ±0.01 1,519 ±4 4.4% 2026-08-10 xhigh
LMArena text/longer_query 1,510 ±2.94 1,503 ±3 4.4% 2026-10-02 high
LMArena text/multi_turn 1,505 ±4.19 1,494 ±4 4.3% 2026-10-02 high
Epoch AI gpqa_diamond outlier decay ×0.26 0.86 ±0.02 1,439 ±12 0.7% 2026-08-06 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 1,459 [1,456–1,462] Percentile 96 · #13 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,625 ±3.27 1,485 ±1 46.0% 2026-10-01 high
LMArena text/coding 1,520 ±3.45 1,488 ±3 44.3% 2026-10-02 high

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 Agents 1,470 [1,464–1,476] Percentile 88 · #8 · components: 2
Source Component Value Elo-eq Weight Measured Variant
LMArena agent/overall 0.08 ±0.01 1,495 ±3 66.2% 2026-10-02 high
τ²-bench (Sierra) tau2/banking_knowledge decay ×0.76 0.40 ±0.05 1,480 ±8 26.7% 2026-07-23 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

S Math 1,489 [1,482–1,496] Percentile 99 · #3 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,518 ±7.03 1,506 ±7 48.4% 2026-10-02 high
Epoch AI frontiermath_t1_3 decay ×0.86 0.87 ±0.02 1,510 ±4 30.9% 2026-06-09 max
Epoch AI frontiermath_t4 decay ×0.86 0.90 ±0.05 1,536 ±7 13.5% 2026-06-09 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 Vision 1,483 [1,478–1,487] Percentile 100 · #1 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,325 ±3.71 1,497 ±3 63.2% 2026-10-02 high
LMArena vision/ocr 1,337 ±3.88 1,507 ±4 15.6% 2026-10-02 high
LMArena vision/diagram 1,356 ±5.44 1,513 ±5 15.0% 2026-10-02 high

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 1,456 [1,448–1,464] Percentile 88 · #5 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena search/overall 1,230 ±4.04 1,473 ±4 93.7% 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 1,499 [1,493–1,504] Percentile 98 · #5 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,495 ±3.91 1,526 ±4 61.8% 2026-10-02 high
LMArena text/longer_query 1,510 ±2.94 1,503 ±3 32.0% 2026-10-02 high

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 1,470 [1,462–1,479] Percentile 97 · #8 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,544 ±5.98 1,489 ±5 93.6% 2026-10-02 high

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 1,454 [1,434–1,474] Percentile 94 · #9 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,488 ±13 1,478 ±11 91.3% 2026-10-02 high

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 1,475 [1,470–1,481] Percentile 98 · #7 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,500 ±2.88 1,493 ±3 93.9% 2026-10-02 high

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 1,469 [1,454–1,485] Percentile 99 · #3 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,500 ±9.27 1,491 ±9 92.3% 2026-10-02 high

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 1,463 [1,448–1,479] Percentile 96 · #6 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,507 ±9.26 1,485 ±8 92.4% 2026-10-02 high

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 1,483 [1,464–1,501] Percentile 98 · #3 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,510 ±14 1,509 ±10 91.5% 2026-10-02 high

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 1,478 [1,462–1,495] Percentile 99 · #2 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,488 ±12 1,502 ±9 92.1% 2026-10-02 high

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 1,461 [1,442–1,480] Percentile 92 · #12 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,492 ±12 1,485 ±11 91.5% 2026-10-02 high

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 1,483 [1,475–1,491] Percentile 98 · #6 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,506 ±4.90 1,502 ±4 93.6% 2026-10-02 high

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
Claude Fable 5 1,484 S

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: Claude Fable 5

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