Skip to content
fedi.software

phi-3-small-8k-instruct in fedi-index

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

fedi-index · General

D 1,126 ±3

Percentile 15 · #249 · provisional

People prefer / Solves tasks

1,121 / —

Human votes vs benchmarks, Elo-eq

Value

-219

$0.2625 per 1M tokens, median of 1 providers

On your hardware

Closed weights: cloud only.

Score decomposition

D General 1,126 [1,123–1,130] Percentile 15 · #249 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,110 ±2.98 1,110 ±3 26.7% 2026-10-02 —
LMArena text/hard_prompts 1,100 ±4.81 1,124 ±4 25.6% 2026-10-02 —
LMArena text/instruction_following 1,087 ±4.39 1,116 ±4 12.9% 2026-10-02 —
LMArena text/expert 1,067 ±9.74 1,122 ±8 10.8% 2026-10-02 —
LMArena text/multi_turn 1,068 ±6.21 1,098 ±6 6.1% 2026-10-02 —
LMArena text/longer_query 1,088 ±6.71 1,109 ±6 5.9% 2026-10-02 —

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 Coding 1,124 [1,115–1,134] Percentile 13 · #243 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,101 ±5.88 1,115 ±5 89.5% 2026-10-02 —

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 Math 1,160 [1,150–1,171] Percentile 17 · #211 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,151 ±6.48 1,154 ±6 87.2% 2026-10-02 —

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 Writing 1,113 [1,104–1,122] Percentile 11 · #238 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,083 ±6.59 1,104 ±7 61.6% 2026-10-02 —
LMArena text/longer_query 1,088 ±6.71 1,109 ±6 31.5% 2026-10-02 —

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 1,111 [1,100–1,121] Percentile 8 · #206 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,061 ±7.16 1,104 ±6 93.3% 2026-10-02 —

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: Deutsch 1,132 [1,113–1,152] Percentile 8 · #132 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,080 ±12 1,126 ±11 91.4% 2026-10-02 —

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: English 1,129 [1,122–1,137] Percentile 17 · #227 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,150 ±4.01 1,124 ±4 93.7% 2026-10-02 —

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: 日本語 1,112 [1,090–1,133] Percentile 2 · #125 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 966 ±16 1,102 ±12 90.7% 2026-10-02 —

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: 한국어 1,063 [1,045–1,081] Percentile 0 · #136 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 893 ±13 1,050 ±10 91.8% 2026-10-02 —

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 1,150 [1,139–1,161] Percentile 11 · #209 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,111 ±6.70 1,147 ±6 93.2% 2026-10-02 —

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
phi-3-small-8k-instruct 1,126 D

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: phi-3-small-8k-instruct

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