---
title: SenseNova-U1.5-8B-MoT-8step-v2-4bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit"
md_url: "https://www.modelscope.cn/models/mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit.md"
repository: mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit
last_updated: 2026-10-12
license: apache-2.0
model_type:
  - neo_chat
architectures:
  - NEOChatModel
base_model:
  - sensenova/SenseNova-U1.5-8B-MoT
base_model_relation: finetune
parameters: 295.5M
tensor_type:
  - U32
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 36
stars: 0
tags:
  - mlx
  - mlx-swift
  - image-generation
  - image-editing
  - any-to-any
  - "native multimodal"
  - distilled
---

# SenseNova-U1.5-8B-MoT-8step-v2-4bit

> SenseNova-U1.5-8B-MoT-8step-v2-4bit - mlx-community 在 ModelScope 开源的模型。SenseNova-U1.5-8B-MoT-8step-v2-4bit

mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit 是 ModelScope 魔搭社区上的 295.5M 参数机器学习模型，采用 apache-2.0 许可，基于 sensenova/SenseNova-U1.5-8B-MoT 构建。

- **Repository**: mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit
- **License**: apache-2.0
- **Parameters**: 295.5M
- **Base model**: sensenova/SenseNova-U1.5-8B-MoT
- **Tags**: mlx, mlx-swift, image-generation, image-editing, any-to-any, native multimodal, distilled
- **Downloads**: 36
- **Stars**: 0
- **Last updated**: 2026-10-12

Source: https://www.modelscope.cn/models/mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit

---

# SenseNova-U1.5-8B-MoT-8step-v2-4bit

Apple Silicon (MLX) artifact of **[SenseNova-U1.5-8B-MoT](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT)** — SenseTime's unified T2I + editing + VQA flagship (NEO-unify Mixture-of-Transformers, pixel-space rectified flow, no VAE).

This artifact is the official **8-step distillation LoRA, V2** ([SenseNova-U1.5-8B-MoT-LoRA-8step-V2](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT-LoRAs/blob/main/SenseNova-U1.5-8B-MoT-LoRA-8step-V2.safetensors), released 2026-09-24, mirrored at [xocialize/SenseNova-U1.5-8B-MoT-LoRAs](https://huggingface.co/xocialize/SenseNova-U1.5-8B-MoT-LoRAs)), α/rank-merged into the generation stream, plus **4-bit (group 64)** stream quantization.
Quantization: group size 64, 4-bit (mlx convention `quantization` block in config.json).

**V2 vs V1.** Upstream describes V2 as its recommended 8-step adapter: better colour balance, more natural contrast and less oversharpening than the V1 adapter in [SenseNova-U1.5-8B-MoT-8step-4bit](https://huggingface.co/mlx-community/SenseNova-U1.5-8B-MoT-8step-4bit). Measured here at fixed seed over five prompts (1024², 8 steps, cfg 1): mean saturation −0.071, RMS contrast −0.049, 99th-pct gradient −30 %, shadow clipping 8.5 pp lower, highlight clipping 14.8 pp lower; composition is preserved (cos(V1, V2) 0.80–0.96). int4 sits at cos 0.976 / 24.2 dB against the V2 bf16 merge-at-load render (tighter than the V1 tier's ~0.92). The V1 artifact stays published unchanged so existing seeds keep reproducing.

**Runtime:** [`sensenova-u1-swift`](https://github.com/xocialize/sensenova-u1-swift) (MLX-Swift), variant `8step-v2-4bit`. Weights are stored in the runtime's key layout (NHWC convs; quantized linears as `weight`/`scales`/`biases`) and load with **zero conversion transient** — peak load memory ≈ resident.

## Performance (M5 Max)

**peak 14.7 GB · 1024² in 3.2 s**

**Use it for:** the smallest, fastest tier for T2I and editing — fits 16 GB-class Macs. Note: 4-bit weight quantization on a diffusion path measurably perturbs the denoise trajectory: at a fixed seed it produces a *different but equally valid* image. Prefer the 8-bit tier when bf16-faithful output matters.

**Editing and VQA run on this tier too** (runtime ≥ 0.5.1, 2026-10-11): the distill adapter touches only the 294 generation-stream projections, so VQA uses the base tier's understanding stream unchanged (answers match token for token), and the 8-step cfg-free edit matches the 50-step base edit (a 2048² jacket edit in 15 s vs 169 s, 28.7 dB apart). Think mode stays a base-tier feature. Evidence: `Docs/receipts/8step-v2-ab/` and AB-R-0490 in the runtime repo.

## Quick start

```bash
git clone https://github.com/xocialize/sensenova-u1-swift && cd sensenova-u1-swift && swift build -c release
hf download mlx-community/SenseNova-U1.5-8B-MoT-8step-v2-4bit --local-dir SenseNova-U1.5-8B-MoT-8step-v2-4bit
.build/release/sensenova-cli --weights SenseNova-U1.5-8B-MoT-8step-v2-4bit \
  --prompt "A cinematic mountain lake at sunrise, realistic photography." \
  --width 1024 --height 1024 --steps 8 --cfg 1.0 --out out.npy
```

The runtime also does instruction editing (`--edit-image`) and VQA (`--vqa`) on this tier, and `<think>` reasoning mode (`--think`) on the base tiers — see the [repository README](https://github.com/xocialize/sensenova-u1-swift) for the full surface and the parity report (component parity < 1e-4; e2e per-pass cosine 0.999+ vs the reference PyTorch implementation).

## Provenance & license

- **Upstream:** [sensenova/SenseNova-U1.5-8B-MoT](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT) @ [`07d76f6`](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT/commit/07d76f61474b9c6e6999e22c559314d3439b8c81), Apache-2.0, by [SenseTime / SenseNova](https://huggingface.co/sensenova) — [paper](https://arxiv.org/abs/2605.12500) · [reference implementation](https://github.com/OpenSenseNova/SenseNova-U1).
- **Adapter:** [sensenova/SenseNova-U1.5-8B-MoT-LoRAs](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT-LoRAs) @ [`f33b8fe`](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT-LoRAs/commit/f33b8fe0216e6f10e5aed58ebbfdd05f02826732), file `SenseNova-U1.5-8B-MoT-LoRA-8step-V2.safetensors` (sha256 `f4190cabd01fa261a45be853275c483d3a8069295c52cf01f7b2b31f4356f201`), Apache-2.0. Merged in fp32 as W += (α/r)·BA over the 294 gen-stream projections and cast; against the V1 artifact exactly 882 tensors differ (294 × weight/scales/biases) and nothing else.
- This repository is a format conversion (official-LoRA merge + bf16 cast + 4-bit group-64 quantization) of the upstream checkpoint, redistributed under the same **Apache-2.0** license with modifications noted here. All credit for the model to the SenseNova team.
- `tokenizer.json` is generated from the upstream `vocab.json`/`merges.txt` (byte-identical tokenization, verified against reference ids).