Grok 4.20 (Reasoning) in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
B 1,414 ±2
Percentile 76 · #70
People prefer / Solves tasks
1,426 / 1,374
Human votes vs benchmarks, Elo-eq
Value
+26
$1.56 per 1M tokens, median of 13 providers · vendor $1.56
On your hardware
Closed weights: cloud only.
Score decomposition
B General 1,414 [1,412–1,417] Percentile 76 · #70 · components: 8
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/overall
|
1,451 ±1.87 | 1,451 ±2 | 21.0% | 2026-10-02 | — |
| LMArena |
text/hard_prompts
|
1,451 ±2.25 | 1,441 ±2 | 21.0% | 2026-10-02 | — |
| Epoch AI |
simpleqa_verified
outlier decay ×0.61 |
0.30 ±0.01 | 1,392 ±5 | 10.4% | 2026-08-27 | — |
| LMArena |
text/instruction_following
|
1,420 ±2.71 | 1,432 ±3 | 10.4% | 2026-10-02 | — |
| LMArena |
text/expert
|
1,440 ±4.12 | 1,427 ±3 | 10.2% | 2026-10-02 | — |
| Epoch AI |
gpqa_diamond
decay ×0.26 |
0.89 ±0.02 | 1,455 ±9 | 7.4% | 2026-07-13 | — |
| LMArena |
text/longer_query
|
1,436 ±2.62 | 1,434 ±2 | 5.2% | 2026-10-02 | — |
| LMArena |
text/multi_turn
|
1,456 ±3.33 | 1,450 ±3 | 5.1% | 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
B Coding 1,402 [1,400–1,405] Percentile 69 · #87 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
webdev/overall
|
1,374 ±3.09 | 1,414 ±1 | 45.8% | 2026-10-01 | — |
| LMArena |
text/coding
|
1,459 ±2.93 | 1,434 ±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
A Math 1,417 [1,411–1,424] Percentile 82 · #47 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,455 ±5.33 | 1,446 ±5 | 52.3% | 2026-10-02 | — |
| Epoch AI |
frontiermath_t1_3
decay ×0.86 |
0.45 ±0.03 | 1,418 ±6 | 28.4% | 2026-07-13 | — |
| Epoch AI |
frontiermath_t4
decay ×0.86 |
0.17 ±0.06 | 1,420 ±10 | 12.0% | 2026-07-13 | — |
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 Vision 1,425 [1,421–1,429] Percentile 62 · #44 · components: 3 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
vision/overall
|
1,263 ±3.11 | 1,441 ±3 | 63.0% | 2026-10-02 | — |
| LMArena |
vision/ocr
|
1,266 ±3.11 | 1,439 ±3 | 15.7% | 2026-10-02 | — |
| LMArena |
vision/diagram
|
1,266 ±4.36 | 1,433 ±4 | 15.3% | 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
A Writing 1,437 [1,432–1,441] Percentile 82 · #50 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,432 ±3.44 | 1,461 ±4 | 62.0% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,436 ±2.62 | 1,434 ±2 | 31.9% | 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: 繁體中文 1,424 [1,417–1,431] Percentile 75 · #57 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,481 ±4.99 | 1,439 ±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
B Language: Deutsch 1,420 [1,404–1,435] Percentile 68 · #47 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/german
|
1,441 ±9.76 | 1,437 ±8 | 92.4% | 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
A Language: English 1,433 [1,429–1,438] Percentile 84 · #44 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,457 ±2.47 | 1,448 ±3 | 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
B Language: Español 1,423 [1,410–1,435] Percentile 70 · #46 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,444 ±7.18 | 1,439 ±7 | 93.0% | 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
A Language: Français 1,439 [1,427–1,451] Percentile 81 · #26 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,476 ±7.34 | 1,457 ±7 | 93.0% | 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: 日本語 1,418 [1,402–1,434] Percentile 70 · #39 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/japanese
|
1,413 ±12 | 1,436 ±9 | 92.2% | 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: 한국어 1,430 [1,416–1,444] Percentile 78 · #31 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/korean
|
1,416 ±10 | 1,448 ±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
B Language: Polski 1,433 [1,420–1,446] Percentile 79 · #28 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/polish
|
1,451 ±8.56 | 1,451 ±7 | 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
A Language: Русский 1,443 [1,436–1,449] Percentile 86 · #33 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,458 ±3.91 | 1,459 ±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
The same in a table
| Model | 2026-10 |
|---|---|
| Grok 4.20 (Reasoning) | 1,414 B |
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.