Grok 2 mini (2024-08-13) in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
C 1,263 ±2
Percentile 30 · #208 · provisional
People prefer / Solves tasks
1,266 / —
Human votes vs benchmarks, Elo-eq
Value
—
On your hardware
Closed weights: cloud only.
Score decomposition
C General 1,263 [1,260–1,265] Percentile 30 · #208 · components: 6 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/overall
|
1,281 ±1.89 | 1,281 ±2 | 25.9% | 2026-11-11 | 2024-08-13 |
| LMArena |
text/hard_prompts
|
1,255 ±2.98 | 1,264 ±3 | 25.5% | 2026-11-11 | 2024-08-13 |
| LMArena |
text/instruction_following
|
1,245 ±2.69 | 1,266 ±2 | 12.8% | 2026-11-11 | 2024-08-13 |
| LMArena |
text/expert
|
1,238 ±5.96 | 1,262 ±5 | 11.9% | 2026-11-11 | 2024-08-13 |
| LMArena |
text/multi_turn
|
1,265 ±3.78 | 1,277 ±3 | 6.2% | 2026-11-11 | 2024-08-13 |
| LMArena |
text/longer_query
|
1,266 ±3.92 | 1,275 ±4 | 6.2% | 2026-11-11 | 2024-08-13 |
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
C Coding 1,257 [1,251–1,263] Percentile 28 · #207 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/coding
|
1,268 ±3.65 | 1,264 ±3 | 88.4% | 2026-11-11 | 2024-08-13 |
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
C Math 1,257 [1,250–1,263] Percentile 30 · #179 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,265 ±3.92 | 1,264 ±4 | 88.3% | 2026-11-11 | 2024-08-13 |
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
C Writing 1,265 [1,260–1,271] Percentile 32 · #179 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,241 ±3.92 | 1,266 ±4 | 62.3% | 2026-11-11 | 2024-08-13 |
| LMArena |
text/longer_query
|
1,266 ±3.92 | 1,275 ±4 | 31.5% | 2026-11-11 | 2024-08-13 |
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
C Language: 繁體中文 1,261 [1,253–1,268] Percentile 25 · #164 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,263 ±5.10 | 1,265 ±4 | 93.7% | 2026-11-11 | 2024-08-13 |
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
D Language: Deutsch 1,286 [1,272–1,300] Percentile 23 · #112 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/german
|
1,273 ±8.82 | 1,292 ±8 | 92.7% | 2026-11-11 | 2024-08-13 |
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
C Language: English 1,276 [1,271–1,280] Percentile 34 · #177 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,298 ±2.46 | 1,280 ±3 | 94.0% | 2026-11-11 | 2024-08-13 |
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
D Language: Español 1,278 [1,257–1,300] Percentile 10 · #135 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,278 ±13 | 1,286 ±12 | 90.7% | 2026-11-11 | 2024-08-13 |
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
D Language: Français 1,274 [1,250–1,298] Percentile 12 · #125 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,286 ±15 | 1,282 ±14 | 89.8% | 2026-11-11 | 2024-08-13 |
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
D Language: 日本語 1,281 [1,269–1,294] Percentile 17 · #106 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/japanese
|
1,213 ±9.20 | 1,287 ±7 | 93.0% | 2026-11-11 | 2024-08-13 |
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
D Language: 한국어 1,274 [1,257–1,290] Percentile 14 · #118 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/korean
|
1,195 ±12 | 1,280 ±9 | 92.1% | 2026-11-11 | 2024-08-13 |
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
C Language: Русский 1,277 [1,270–1,284] Percentile 26 · #171 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,261 ±4.06 | 1,282 ±4 | 93.8% | 2026-11-11 | 2024-08-13 |
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
History
The same in a table
| Model | 2026-11 |
|---|---|
| Grok 2 mini (2024-08-13) | 1,263 C |
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.