Skip to content
fedi.software

DeepSeek V4 Flash 0423 in fedi-index

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

fedi-index · General

B 1,407 ±7

Percentile 73 · #81

People prefer / Solves tasks

1,419 / 1,392

Human votes vs benchmarks, Elo-eq

Value

+71

$0.175 per 1M tokens, median of 39 providers

On your hardware

  • 128 GB unified memory #1 · UD-IQ3_XXS · ≈26 tok/s

Score decomposition

B General 1,407 [1,400–1,413] Percentile 73 · #81 · components: 7
Source Component Value Elo-eq Weight Measured Variant
Epoch AI eci 146 ±1.01 1,407 ±8 41.6% 2026-04-24 —
LMArena text/overall 1,432 ±2.06 1,432 ±2 15.0% 2026-10-02 —
LMArena text/hard_prompts 1,443 ±2.42 1,434 ±2 14.9% 2026-10-02 —
LMArena text/instruction_following 1,421 ±2.96 1,433 ±3 7.4% 2026-10-02 —
LMArena text/expert 1,441 ±4.43 1,428 ±4 7.2% 2026-10-02 —
LMArena text/longer_query 1,434 ±2.82 1,433 ±3 3.7% 2026-10-02 —
LMArena text/multi_turn 1,449 ±3.70 1,443 ±3 3.6% 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

A Coding 1,428 [1,426–1,431] Percentile 82 · #51 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,581 ±4.71 1,473 ±1 45.8% 2026-10-01 high
LMArena text/coding 1,457 ±3.15 1,432 ±3 44.6% 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

B Math 1,392 [1,382–1,402] Percentile 63 · #94 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,427 ±6.01 1,419 ±6 87.4% 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

B Writing 1,418 [1,413–1,423] Percentile 74 · #71 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,403 ±3.83 1,432 ±4 61.8% 2026-10-02 —
LMArena text/longer_query 1,434 ±2.82 1,433 ±3 32.0% 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

B Language: 繁體中文 1,413 [1,405–1,422] Percentile 69 · #70 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,468 ±5.82 1,428 ±5 93.6% 2026-10-02 high

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: Deutsch 1,401 [1,384–1,418] Percentile 57 · #62 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,418 ±11 1,418 ±9 92.0% 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

B Language: English 1,417 [1,412–1,422] Percentile 71 · #79 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,440 ±2.69 1,430 ±3 93.9% 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

B Language: Español 1,415 [1,401–1,429] Percentile 60 · #60 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,436 ±8.18 1,432 ±8 92.7% 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

C Language: Français 1,407 [1,394–1,419] Percentile 48 · #70 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,438 ±7.63 1,422 ±7 92.9% 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

B Language: 日本語 1,411 [1,394–1,428] Percentile 63 · #48 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,403 ±13 1,429 ±10 92.0% 2026-10-02 high

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: 한국어 1,406 [1,392–1,421] Percentile 58 · #58 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,384 ±11 1,423 ±8 92.5% 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

B Language: Polski 1,421 [1,407–1,436] Percentile 67 · #44 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,437 ±9.51 1,439 ±8 92.6% 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

B Language: Русский 1,417 [1,410–1,424] Percentile 71 · #69 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,428 ±4.32 1,432 ±4 93.8% 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

History

fedi-index by month
The same in a table
Model 2026-10
DeepSeek V4 Flash 0423 1,407 B

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: DeepSeek V4 Flash 0423

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