Skip to content
fedi.software

GPT-6 Sol in fedi-index

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

fedi-index · General

A 1,434 ±8

Percentile 84 · #49

People prefer / Solves tasks

1,397 / 1,494

Human votes vs benchmarks, Elo-eq

Value

+22

$4 per 1M tokens, median of 22 providers · vendor $4

On your hardware

Closed weights: cloud only.

Score decomposition

A General 1,434 [1,426–1,441] Percentile 84 · #49 · components: 9
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 163 ±1.66 1,538 ±13 27.6% 2026-09-22 —
LMArena text/overall 1,395 ±3.54 1,395 ±4 17.5% 2026-10-02 max
LMArena text/hard_prompts 1,419 ±4.34 1,412 ±4 17.3% 2026-10-02 max
LMArena text/instruction_following 1,413 ±5.66 1,425 ±5 8.2% 2026-10-02 max
LMArena text/expert 1,439 ±9.92 1,426 ±8 7.2% 2026-10-02 max
Epoch AI simpleqa_verified decay ×0.61 0.61 ±0.02 1,488 ±5 4.5% 2026-09-22 max
LMArena text/longer_query 1,414 ±5.24 1,413 ±5 4.2% 2026-10-02 max
LMArena text/multi_turn 1,413 ±8.43 1,411 ±8 3.7% 2026-10-02 max
Epoch AI gpqa_diamond decay ×0.26 0.94 ±0.01 1,478 ±7 1.8% 2026-09-22 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 1,444 [1,440–1,448] Percentile 91 · #25 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,689 ±5.84 1,503 ±2 60.7% 2026-10-01 max
LMArena text/coding outlier 1,448 ±6.78 1,424 ±6 26.5% 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 Agents 1,470 [1,460–1,480] Percentile 90 · #7 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena agent/overall 0.10 ±0.01 1,502 ±6 89.3% 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 1,447 [1,436–1,458] Percentile 90 · #26 · components: 3
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,407 ±14 1,400 ±13 38.6% 2026-10-02 max
Epoch AI frontiermath_t1_3 decay ×0.86 0.90 ±0.02 1,516 ±4 37.4% 2026-09-22 max
Epoch AI frontiermath_t4 decay ×0.86 0.90 ±0.05 1,536 ±8 15.3% 2026-09-22 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

B Writing 1,397 [1,388–1,406] Percentile 64 · #98 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,382 ±7.40 1,410 ±8 59.4% 2026-10-02 max
LMArena text/longer_query 1,414 ±5.24 1,413 ±5 33.7% 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

C Language: 繁體中文 1,362 [1,345–1,379] Percentile 42 · #130 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,403 ±12 1,376 ±9 92.0% 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

B Language: English 1,374 [1,364–1,384] Percentile 55 · #123 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,398 ±5.37 1,386 ±6 93.3% 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

C Language: Español 1,361 [1,334–1,387] Percentile 26 · #111 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,381 ±17 1,381 ±15 88.7% 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

B Language: Русский 1,399 [1,385–1,414] Percentile 60 · #93 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,409 ±8.84 1,415 ±8 92.6% 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

History

fedi-index by month
The same in a table
Model 2026-10
GPT-6 Sol 1,434 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: GPT-6 Sol

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