Muse Spark 1.3 in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
S 1,469 ±8
Percentile 97 · #9
People prefer / Solves tasks
1,470 / 1,468
Human votes vs benchmarks, Elo-eq
Value
+74
$2 per 1M tokens, median of 10 providers · vendor $0.125
On your hardware
Closed weights: cloud only.
Score decomposition
S General 1,469 [1,462–1,477] Percentile 97 · #9 · components: 7
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| Epoch AI |
eci
|
157 ±1.16 | 1,491 ±9 | 41.0% | 2026-09-02 | — |
| LMArena |
text/overall
|
1,490 ±3.14 | 1,490 ±3 | 15.4% | 2026-10-02 | max |
| LMArena |
text/hard_prompts
|
1,504 ±3.71 | 1,488 ±3 | 15.3% | 2026-10-02 | max |
| LMArena |
text/instruction_following
|
1,480 ±4.71 | 1,489 ±4 | 7.4% | 2026-10-02 | max |
| LMArena |
text/expert
|
1,518 ±8.10 | 1,491 ±7 | 6.7% | 2026-10-02 | max |
| LMArena |
text/longer_query
|
1,491 ±4.36 | 1,486 ±4 | 3.8% | 2026-10-02 | max |
| LMArena |
text/multi_turn
|
1,486 ±7.13 | 1,477 ±6 | 3.4% | 2026-10-02 | 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 Coding 1,461 [1,457–1,465] Percentile 97 · #10 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
webdev/overall
|
1,657 ±4.36 | 1,494 ±1 | 47.1% | 2026-10-01 | max |
| LMArena |
text/coding
|
1,515 ±5.39 | 1,484 ±5 | 43.0% | 2026-10-02 | 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 Agents 1,449 [1,446–1,451] Percentile 70 · #18 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
agent/overall
|
0.04 ±0.00 | 1,475 ±2 | 90.5% | 2026-10-02 | 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 Math 1,460 [1,448–1,471] Percentile 94 · #16 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,502 ±12 | 1,491 ±12 | 43.8% | 2026-10-02 | max |
| Epoch AI |
frontiermath_t1_3
decay ×0.86 |
0.74 ±0.03 | 1,482 ±6 | 35.5% | 2026-09-16 | xhigh |
| Epoch AI |
frontiermath_t4
decay ×0.86 |
0.46 ±0.08 | 1,466 ±13 | 11.9% | 2026-09-18 | 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 Vision 1,467 [1,460–1,474] Percentile 92 · #10 · components: 3 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
vision/overall
|
1,310 ±5.32 | 1,484 ±5 | 64.7% | 2026-10-02 | max |
| LMArena |
vision/ocr
|
1,319 ±6.11 | 1,490 ±6 | 15.4% | 2026-10-02 | max |
| LMArena |
vision/diagram
|
1,330 ±9.96 | 1,490 ±9 | 13.2% | 2026-10-02 | 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 1,468 [1,460–1,476] Percentile 94 · #17 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,458 ±6.30 | 1,488 ±6 | 60.2% | 2026-10-02 | max |
| LMArena |
text/longer_query
|
1,491 ±4.36 | 1,486 ±4 | 33.2% | 2026-10-02 | 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 1,456 [1,442–1,470] Percentile 92 · #18 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,529 ±9.95 | 1,476 ±8 | 92.6% | 2026-10-02 | 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 1,468 [1,459–1,476] Percentile 97 · #10 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,492 ±4.46 | 1,486 ±5 | 93.6% | 2026-10-02 | 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: Español 1,448 [1,422–1,474] Percentile 91 · #14 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,487 ±16 | 1,479 ±15 | 89.0% | 2026-10-02 | 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: Français 1,469 [1,442–1,496] Percentile 98 · #3 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,528 ±17 | 1,504 ±16 | 88.5% | 2026-10-02 | 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: Русский 1,470 [1,458–1,483] Percentile 95 · #13 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,493 ±7.66 | 1,490 ±7 | 93.0% | 2026-10-02 | 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 | 2026-10 |
|---|---|
| Muse Spark 1.3 | 1,469 S |
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.