OpenAssistant Pythia 12B in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
D 961 ±7
Percentile 1 · #293 · provisional
People prefer / Solves tasks
943 / —
Human votes vs benchmarks, Elo-eq
Value
—
On your hardware
Closed weights: cloud only.
Score decomposition
D General 961 [954–967] Percentile 1 · #293 · components: 4 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/overall
|
916 ±5.44 | 916 ±5 | 34.8% | 2026-11-11 | — |
| LMArena |
text/hard_prompts
|
885 ±9.45 | 929 ±8 | 29.8% | 2026-11-11 | — |
| LMArena |
text/instruction_following
|
887 ±7.92 | 927 ±8 | 15.8% | 2026-11-11 | — |
| LMArena |
text/multi_turn
|
869 ±13 | 917 ±12 | 6.2% | 2026-11-11 | — |
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 Coding 979 [961–997] Percentile 2 · #282 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/coding
|
900 ±12 | 936 ±11 | 83.8% | 2026-11-11 | — |
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 Math 953 [935–970] Percentile 1 · #253 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
892 ±11 | 906 ±11 | 84.2% | 2026-11-11 | — |
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 Writing 963 [945–982] Percentile 2 · #259 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
923 ±10 | 941 ±10 | 91.5% | 2026-11-11 | — |
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: 繁體中文 930 [907–953] Percentile 0 · #219 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
803 ±17 | 900 ±13 | 90.0% | 2026-11-11 | — |
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: English 928 [917–940] Percentile 2 · #264 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
945 ±5.96 | 908 ±6 | 93.2% | 2026-11-11 | — |
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 |
|---|---|
| OpenAssistant Pythia 12B | 961 D |
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.