Skip to content
fedi.software

Claude Sonnet 5.5 in fedi-index

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

fedi-index · General

A 1,462 ±6

Percentile 94 · #17

People prefer / Solves tasks

1,460 / 1,467

Human votes vs benchmarks, Elo-eq

Value

+50

$4 per 1M tokens, median of 20 providers · vendor $4

On your hardware

Closed weights: cloud only.

Score decomposition

A General 1,462 [1,456–1,468] Percentile 94 · #17 · components: 9
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,467 ±5.31 1,467 ±5 21.8% 2026-10-02 xhigh
LMArena text/hard_prompts 1,493 ±6.67 1,478 ±6 21.1% 2026-10-02 xhigh
Epoch AI eci outlier 165 ±1.88 1,556 ±15 16.2% 2026-09-28 —
LMArena text/instruction_following 1,490 ±8.96 1,498 ±8 9.3% 2026-10-02 xhigh
LMArena text/expert 1,550 ±16 1,517 ±13 7.1% 2026-10-02 xhigh
LMArena text/longer_query 1,488 ±8.38 1,483 ±8 4.8% 2026-10-02 xhigh
LMArena text/multi_turn 1,463 ±14 1,456 ±13 3.6% 2026-10-02 xhigh
Epoch AI simpleqa_verified outlier decay ×0.61 0.46 ±0.02 1,443 ±5 2.9% 2026-09-29 max
Epoch AI gpqa_diamond decay ×0.26 0.96 ±0.01 1,484 ±6 2.4% 2026-09-29 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 1,478 [1,471–1,486] Percentile 100 · #2 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,786 ±9.28 1,531 ±3 51.3% 2026-10-01 xhigh
LMArena text/coding 1,519 ±11 1,487 ±10 37.6% 2026-10-02 xhigh

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 1,479 [1,466–1,492] Percentile 95 · #4 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena agent/overall 0.13 ±0.02 1,515 ±8 88.5% 2026-10-02 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 Math 1,470 [1,463–1,477] Percentile 98 · #7 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
Epoch AI frontiermath_t1_3 decay ×0.86 0.89 ±0.02 1,513 ±4 62.3% 2026-09-29 max
Epoch AI frontiermath_t4 decay ×0.86 0.80 ±0.06 1,521 ±10 23.2% 2026-09-29 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 1,446 [1,436–1,456] Percentile 81 · #22 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,290 ±7.72 1,465 ±7 66.4% 2026-10-02 xhigh
LMArena vision/ocr 1,291 ±8.97 1,463 ±9 15.2% 2026-10-02 xhigh
LMArena vision/diagram 1,313 ±16 1,475 ±14 10.9% 2026-10-02 xhigh

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 1,457 [1,443–1,472] Percentile 91 · #26 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,450 ±12 1,480 ±12 55.2% 2026-10-02 xhigh
LMArena text/longer_query 1,488 ±8.38 1,483 ±8 36.2% 2026-10-02 xhigh

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 1,458 [1,442–1,474] Percentile 95 · #14 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,487 ±8.55 1,480 ±9 92.2% 2026-10-02 xhigh

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 1,422 [1,399–1,444] Percentile 74 · #62 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,443 ±14 1,445 ±13 90.3% 2026-10-02 xhigh

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 Sonnet 5.5 1,462 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: Claude Sonnet 5.5

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