Text (LLM), image, video, and multimodal—alongside the combined Model board.
Model leaderboard (all)
Leistung über Aufgaben hinweg (Multimodal/Vision/Sprache; Beispieldaten).
Cross-task overview; multimodal, vision, and language may split into additional columns or child boards once eval JSON is wired.
Public ranking policy: rows are sorted by composite score (desc). Composite score is a weighted sum of normalized sub-metrics; ties are broken by higher recent activity.
| Rang | Modell | Anbieter / Team | Typ | Punktzahl | Hinweise |
|---|---|---|---|---|---|
| 1 | Demo-Vision-Pro | Demo Lab | Multimodal | 94.2 | Ausgewogenes Bild + Text |
| 2 | NorthStar-MM | North AI | Multimodal | 92.8 | Starke Langkontext-Szenarien |
| 3 | Aurora-VL-7B | Aurora | Vision-Sprache | 91.5 | Edge-tauglich |
| 4 | Helix-3 | Helix Research | Allgemein | 90.1 | Stabiler Tool-Aufruf |
| 5 | Kite-Small | Kite | Sprache | 88.6 | Gutes Preis-Leistungs-Verhältnis |
| 6 | Lattice-R1 | Lattice | Reasoning | 87.9 | Starke Teilwerte Mathe/Code |
| 7 | Pulse-Audio-2 | Pulse | Sprach-Multimodal | 86.4 | ASR/TTS kombiniert |
| 8 | Quark-Mini | Quark Systems | Sprache | 85.2 | Geringe Latenz |