Skip to content
fedi.software

Gemini 2.5 Flash Lite in fedi-index

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

fedi-index · General

C 1,351 ±2

Percentile 47 · #155 · provisional

People prefer / Solves tasks

1,360 / —

Human votes vs benchmarks, Elo-eq

Value

+16

$0.175 per 1M tokens, median of 20 providers · vendor $0.175

On your hardware

Closed weights: cloud only.

Score decomposition

C General 1,351 [1,349–1,354] Percentile 47 · #155 · components: 6 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,369 ±2.29 1,369 ±2 26.0% 2026-10-02 thinking
LMArena text/hard_prompts 1,373 ±2.95 1,371 ±3 25.8% 2026-10-02 thinking
LMArena text/instruction_following 1,366 ±3.56 1,381 ±3 12.7% 2026-10-02 thinking
LMArena text/expert 1,365 ±7.49 1,366 ±6 11.4% 2026-10-02 thinking
LMArena text/longer_query 1,372 ±3.98 1,375 ±4 6.3% 2026-10-02 thinking
LMArena text/multi_turn 1,360 ±4.25 1,362 ±4 6.2% 2026-10-02 thinking

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 1,342 [1,335–1,348] Percentile 41 · #165 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,373 ±4.05 1,357 ±4 90.1% 2026-10-02 thinking

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 1,337 [1,327–1,348] Percentile 42 · #147 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,362 ±6.38 1,357 ±6 87.3% 2026-10-02 thinking

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 Vision 1,359 [1,354–1,364] Percentile 32 · #77 · components: 3 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena vision/overall 1,187 ±3.78 1,373 ±3 63.9% 2026-10-02 thinking
LMArena vision/ocr 1,188 ±4.69 1,363 ±4 15.4% 2026-10-02 thinking
LMArena vision/diagram 1,186 ±6.84 1,362 ±6 14.4% 2026-10-02 thinking

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 1,376 [1,369–1,382] Percentile 56 · #119 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,366 ±4.76 1,394 ±5 61.6% 2026-10-02 thinking
LMArena text/longer_query 1,372 ±3.98 1,375 ±4 32.1% 2026-10-02 thinking

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 1,364 [1,354–1,373] Percentile 43 · #128 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,401 ±6.42 1,375 ±5 93.5% 2026-10-02 thinking

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: Deutsch 1,365 [1,348–1,381] Percentile 38 · #90 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,372 ±11 1,378 ±9 92.1% 2026-10-02 thinking

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 1,349 [1,344–1,355] Percentile 45 · #150 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,372 ±2.96 1,359 ±3 93.9% 2026-10-02 thinking

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: Español 1,351 [1,333–1,369] Percentile 23 · #115 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,363 ±11 1,364 ±10 91.7% 2026-10-02 thinking

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: Français 1,356 [1,332–1,380] Percentile 25 · #102 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,385 ±15 1,373 ±14 89.8% 2026-10-02 thinking

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: 日本語 1,374 [1,359–1,388] Percentile 39 · #78 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,348 ±11 1,388 ±8 92.5% 2026-10-02 thinking

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: 한국어 1,389 [1,373–1,404] Percentile 46 · #74 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,359 ±12 1,404 ±9 92.3% 2026-10-02 thinking

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: Polski 1,364 [1,355–1,373] Percentile 23 · #101 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,362 ±5.91 1,375 ±5 93.5% 2026-10-02 thinking

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: Русский 1,362 [1,352–1,373] Percentile 43 · #133 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,363 ±6.13 1,374 ±6 93.4% 2026-10-02 thinking

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
Gemini 2.5 Flash Lite 1,351 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: Gemini 2.5 Flash Lite

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