Kimi K3.1 Pro in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
A 1,470 ±9
Percentile 94 · #19
People prefer / Solves tasks
1,444 / 1,500
Human votes vs benchmarks, Elo-eq
Value
+53
$4 per 1M tokens, median of 56 providers · vendor $4
On your hardware
Does not fit the reference setups.
Score decomposition
A General 1,470 [1,461–1,478] Percentile 94 · #19 · components: 6
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| Epoch AI |
eci
|
161 ±1.28 | 1,528 ±10 | 41.7% | 2026-10-23 | — |
| LMArena |
text/overall
|
1,459 ±5.09 | 1,459 ±5 | 18.1% | 2026-11-11 | max |
| LMArena |
text/hard_prompts
|
1,471 ±6.36 | 1,458 ±6 | 17.6% | 2026-11-11 | max |
| LMArena |
text/instruction_following
|
1,472 ±8.48 | 1,482 ±8 | 7.8% | 2026-11-11 | max |
| LMArena |
text/longer_query
|
1,468 ±7.85 | 1,464 ±7 | 4.1% | 2026-11-11 | max |
| LMArena |
text/multi_turn
|
1,449 ±13 | 1,443 ±12 | 3.2% | 2026-11-11 | max |
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
S Coding 1,463 [1,458–1,469] Percentile 95 · #15 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
webdev/overall
|
1,686 ±4.84 | 1,503 ±1 | 37.5% | 2026-11-15 | max |
| LMArena |
text/coding
|
1,480 ±10 | 1,453 ±9 | 27.7% | 2026-11-11 | max |
| Terminal-Bench |
tbench/terminal-bench
decay ×0.90 |
0.60 ±0.02 | 1,513 ±4 | 25.3% | 2026-10-24 | max |
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 Agents 1,460 [1,456–1,464] Percentile 82 · #15 · components: 2
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
agent/overall
|
0.04 ±0.01 | 1,474 ±3 | 67.8% | 2026-11-15 | max |
| Terminal-Bench |
tbench/terminal-bench
decay ×0.90 |
0.60 ±0.02 | 1,513 ±4 | 23.4% | 2026-10-24 | max |
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 Writing 1,445 [1,432–1,458] Percentile 83 · #46 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,439 ±11 | 1,468 ±11 | 56.9% | 2026-11-11 | max |
| LMArena |
text/longer_query
|
1,468 ±7.85 | 1,464 ±7 | 35.0% | 2026-11-11 | max |
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 Language: 繁體中文 1,439 [1,416–1,463] Percentile 81 · #42 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,516 ±17 | 1,466 ±13 | 89.9% | 2026-11-11 | max |
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 Language: English 1,438 [1,422–1,454] Percentile 86 · #39 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,466 ±8.58 | 1,458 ±9 | 92.1% | 2026-11-11 | max |
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 1,433 [1,414–1,451] Percentile 78 · #51 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,452 ±12 | 1,454 ±10 | 91.5% | 2026-11-11 | max |
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
The same in a table
| Model | 2026-11 |
|---|---|
| Kimi K3.1 Pro | 1,470 A |
Badge
Show the model's fedi-index on your site or in a README. The badge updates itself; it links to this page.
Data CC BY 4.0: keep the link to fedi.software.