Skip to content
fedi.software

olmo-3-32b-think in fedi-index

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

fedi-index · General

C 1,287 ±5

Percentile 32 · #199 · provisional

People prefer / Solves tasks

1,294 / —

Human votes vs benchmarks, Elo-eq

Value

—

On your hardware

Closed weights: cloud only.

Score decomposition

C General 1,287 [1,283–1,292] Percentile 32 · #199 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,300 ±4.21 1,300 ±4 28.2% 2026-10-02 think
LMArena text/hard_prompts 1,302 ±5.65 1,306 ±5 27.3% 2026-10-02 think
LMArena text/instruction_following 1,275 ±8.01 1,295 ±8 12.1% 2026-10-02 think
LMArena text/expert 1,272 ±18 1,290 ±15 7.7% 2026-10-02 think
LMArena text/longer_query 1,297 ±8.20 1,304 ±8 6.0% 2026-10-02 think
LMArena text/multi_turn 1,290 ±11 1,299 ±10 5.2% 2026-10-02 think

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 Coding 1,296 [1,282–1,310] Percentile 33 · #187 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,319 ±9.38 1,309 ±8 88.0% 2026-10-02 think

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 Math 1,289 [1,265–1,314] Percentile 34 · #168 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,315 ±16 1,312 ±15 79.5% 2026-10-02 think

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 Writing 1,282 [1,269–1,296] Percentile 33 · #180 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,255 ±11 1,281 ±11 57.4% 2026-10-02 think
LMArena text/longer_query 1,297 ±8.20 1,304 ±8 34.5% 2026-10-02 think

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 1,285 [1,265–1,305] Percentile 26 · #166 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,298 ±14 1,293 ±11 91.0% 2026-10-02 think

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: English 1,324 [1,312–1,335] Percentile 39 · #168 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,347 ±6.07 1,332 ±6 93.1% 2026-10-02 think

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

D Language: Русский 1,270 [1,250–1,290] Percentile 24 · #178 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,255 ±13 1,277 ±11 91.1% 2026-10-02 think

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
olmo-3-32b-think 1,287 C

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: olmo-3-32b-think

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