Back to Model
T

TenEros Max Learned Quantizations

Model Nov 4, 2026

LokkenJP's community activation-guided quantizations of TenStrip's MiniMax H3-based 10Eros Max video model, including Beta7 fix1 HQ and an older compact checkpoint.

Visit primary link 2 sources attached
Author table comparing Beta7 Test3 and fix1 quantization reconstruction errors across matrix families.
LokkenJP’s calibration reconstruction table. These are the author’s tensor-error measurements, not a controlled video-quality or inference-speed benchmark; sources, captures and objectives differ. Source: LokkenJP
multimodalself-hosted

What happened lately

TenEros Max Learned Quantizations timeline

The author recommends an 18.73 GB Beta7 fix1 HQ transformer using W6 and INT8 weights. The separate 14.65 GB compact checkpoint uses the original Beta7 source and does not include fix1 changes. These are disk sizes, not complete-pipeline VRAM requirements.

The project reports mechanical checks and subjective ComfyUI acceptance dated November 2, 2026. It does not establish controlled video-quality or speed improvements. The upstream model and derivative carry community-license terms and a Not-For-All-Audiences designation.