Hy3 in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
B 1,411 ±3
Percentile 74 · #78 · provisional
People prefer / Solves tasks
1,424 / —
Human votes vs benchmarks, Elo-eq
Value
+68
$0.231 per 1M tokens, median of 13 providers
On your hardware
- GPU 24 GB #26 · IQ1_M · strong compression · ≈12 tok/s
- 128 GB unified memory #37 · Q2_K_L · strong compression · ≈14 tok/s
Score decomposition
B General 1,411 [1,407–1,414] Percentile 74 · #78 · components: 6 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/overall
|
1,441 ±3.32 | 1,441 ±3 | 26.2% | 2026-10-02 | — |
| LMArena |
text/hard_prompts
|
1,449 ±3.92 | 1,439 ±4 | 26.1% | 2026-10-02 | — |
| LMArena |
text/instruction_following
|
1,425 ±5.06 | 1,437 ±5 | 12.5% | 2026-10-02 | — |
| LMArena |
text/expert
|
1,460 ±8.93 | 1,443 ±7 | 11.1% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,442 ±4.62 | 1,439 ±4 | 6.4% | 2026-10-02 | — |
| LMArena |
text/multi_turn
|
1,435 ±7.50 | 1,431 ±7 | 5.7% | 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 Coding 1,421 [1,417–1,426] Percentile 78 · #61 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
webdev/overall
|
1,507 ±4.42 | 1,452 ±1 | 47.3% | 2026-10-01 | — |
| LMArena |
text/coding
|
1,464 ±5.66 | 1,439 ±5 | 42.8% | 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
C Agents 1,408 [1,402–1,413] Percentile 28 · #42 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
agent/overall
|
-0.05 ±0.01 | 1,430 ±3 | 90.2% | 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
A Math 1,418 [1,399–1,438] Percentile 83 · #45 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,473 ±13 | 1,463 ±12 | 82.8% | 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
B Writing 1,418 [1,409–1,427] Percentile 74 · #70 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,402 ±6.92 | 1,430 ±7 | 59.6% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,442 ±4.62 | 1,439 ±4 | 33.6% | 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
B Language: 繁體中文 1,430 [1,415–1,446] Percentile 78 · #51 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,495 ±11 | 1,449 ±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
B Language: English 1,425 [1,416–1,435] Percentile 77 · #63 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,450 ±4.82 | 1,441 ±5 | 93.5% | 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 1,421 [1,397–1,446] Percentile 66 · #51 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,453 ±15 | 1,447 ±14 | 89.4% | 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
B Language: Français 1,414 [1,389–1,440] Percentile 56 · #60 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,458 ±16 | 1,440 ±14 | 89.2% | 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: Русский 1,419 [1,405–1,433] Percentile 73 · #64 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,433 ±8.55 | 1,436 ±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
History
The same in a table
| Model | 2026-10 |
|---|---|
| Hy3 | 1,411 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.