Cyclone-Labs/Eclipsed-Phoenix-26B-A4B

Image-Text-to-Text • 26B • Updated 4 days ago • 40 • 6

Note New and shiny.

TheDrummer/Orion-26B-A4B-v1

26B • Updated 3 days ago • 422 • 26

Note This one hits different. Few issues here and there, needs some proper merge or something. Good first impression overall.

Koshkasa/Vortex5_Shadow-Siren-26B-A4B-APEX-GGUF

Text Generation • 25B • Updated about 1 month ago • 1.14k • 6

Note Slightly lower performance on my hardware (26 t/s vs 20 t/s) compared to IQ4_XS, but it seems like it's worth it.

Vortex5/Chimera-X-26B-A4B

Image-Text-to-Text • 27B • Updated Jul 16 • 89 • 13

Note What I've learned so far: this is still the most stable eRP model I've ever tried. Never heretic or ablate MoE models - they're too fragile for this. A correct system prompt and PHI are critical for the best Gemma 4 rabbit-hole experience.