Skip to content
fedi.software

kimi-k2-thinking-turbo in fedi-index

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

fedi-index · General

B 1,399 ±3

Percentile 68 · #96

People prefer / Solves tasks

1,405 / 1,374

Human votes vs benchmarks, Elo-eq

Value

-5

$2.86 per 1M tokens, median of 1 providers

On your hardware

Closed weights: cloud only.

Score decomposition

B General 1,399 [1,396–1,402] Percentile 68 · #96 · components: 7
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,414 ±1.57 1,414 ±2 21.9% 2026-10-02 —
LMArena text/hard_prompts 1,428 ±1.98 1,420 ±2 21.9% 2026-10-02 —
Epoch AI gpqa_diamond decay ×0.26 0.84 ±0.02 1,432 ±10 14.5% 2025-11-11 —
LMArena text/instruction_following 1,403 ±2.52 1,416 ±2 10.8% 2026-10-02 —
LMArena text/expert 1,439 ±4.53 1,426 ±4 10.4% 2026-10-02 —
LMArena text/longer_query 1,415 ±2.53 1,415 ±2 5.4% 2026-10-02 —
LMArena text/multi_turn 1,414 ±3.16 1,412 ±3 5.4% 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 Coding 1,394 [1,391–1,396] Percentile 65 · #98 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena webdev/overall 1,321 ±3.78 1,399 ±1 45.7% 2026-10-01 —
LMArena text/coding 1,453 ±2.75 1,429 ±2 44.7% 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

B Math 1,395 [1,387–1,403] Percentile 66 · #87 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,429 ±4.87 1,421 ±5 87.9% 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 Writing 1,393 [1,389–1,398] Percentile 62 · #101 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,373 ±3.35 1,401 ±3 62.0% 2026-10-02 —
LMArena text/longer_query 1,415 ±2.53 1,415 ±2 31.8% 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

B Language: 繁體中文 1,406 [1,399–1,412] Percentile 61 · #87 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 1,457 ±4.58 1,419 ±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

C Language: Deutsch 1,379 [1,365–1,394] Percentile 46 · #79 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 1,390 ±9.14 1,394 ±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: English 1,403 [1,399–1,407] Percentile 64 · #99 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,426 ±2.21 1,416 ±2 94.0% 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

C Language: Español 1,391 [1,379–1,403] Percentile 42 · #87 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/spanish 1,407 ±7.11 1,405 ±7 93.1% 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 1,395 [1,381–1,409] Percentile 42 · #78 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/french 1,425 ±8.16 1,410 ±8 92.8% 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

C Language: 日本語 1,378 [1,360–1,396] Percentile 42 · #75 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/japanese 1,356 ±13 1,394 ±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

C Language: 한국어 1,369 [1,355–1,382] Percentile 34 · #90 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/korean 1,329 ±9.94 1,382 ±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: Polski 1,390 [1,379–1,401] Percentile 38 · #82 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/polish 1,396 ±7.19 1,404 ±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

B Language: Русский 1,386 [1,380–1,392] Percentile 53 · #110 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 1,390 ±3.58 1,398 ±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

History

fedi-index by month
The same in a table
Model 2026-10
kimi-k2-thinking-turbo 1,399 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: kimi-k2-thinking-turbo

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