Skip to content
fedi.software

Granite-4.0-H-Small in fedi-index

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

fedi-index · General

C 1,243 ±5

Percentile 28 · #215 · provisional

People prefer / Solves tasks

1,246 / —

Human votes vs benchmarks, Elo-eq

Value

-62

$0.1075 per 1M tokens, median of 1 providers

On your hardware

Closed weights: cloud only.

Score decomposition

C General 1,243 [1,238–1,248] Percentile 28 · #215 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,241 ±4.31 1,242 ±4 28.1% 2026-11-11 —
LMArena text/hard_prompts 1,241 ±5.77 1,251 ±5 27.1% 2026-11-11 —
LMArena text/instruction_following 1,222 ±7.75 1,245 ±7 12.2% 2026-11-11 —
LMArena text/expert 1,251 ±17 1,272 ±14 8.0% 2026-11-11 —
LMArena text/longer_query 1,242 ±8.64 1,253 ±8 5.9% 2026-11-11 —
LMArena text/multi_turn 1,243 ±10 1,257 ±9 5.5% 2026-11-11 —

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 1,241 [1,228–1,254] Percentile 26 · #213 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,249 ±8.71 1,247 ±8 86.3% 2026-11-11 —

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 1,242 [1,218–1,266] Percentile 29 · #183 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,253 ±16 1,252 ±16 79.4% 2026-11-11 —

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 1,238 [1,224–1,253] Percentile 29 · #188 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,210 ±12 1,234 ±12 55.7% 2026-11-11 —
LMArena text/longer_query 1,242 ±8.64 1,253 ±8 35.7% 2026-11-11 —

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: 繁體中文 1,250 [1,227–1,273] Percentile 25 · #165 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,251 ±16 1,256 ±13 90.2% 2026-11-11 —

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 1,244 [1,232–1,256] Percentile 30 · #189 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,266 ±6.30 1,247 ±7 93.0% 2026-11-11 —

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 1,225 [1,202–1,249] Percentile 16 · #194 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,201 ±15 1,228 ±13 90.0% 2026-11-11 —

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 2026-11
Granite-4.0-H-Small 1,243 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: Granite-4.0-H-Small

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