---
title: Inkling-NVFP4-mlx-4bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/Inkling-NVFP4-mlx-4bit"
md_url: "https://www.modelscope.cn/models/mlx-community/Inkling-NVFP4-mlx-4bit.md"
repository: mlx-community/Inkling-NVFP4-mlx-4bit
last_updated: 2026-07-17
license: apache-2.0
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - inkling_mm_model
architectures:
  - InklingForConditionalGeneration
base_model:
  - thinkingmachines/Inkling-NVFP4
base_model_relation: finetune
parameters: 176.5B
tensor_type:
  - BF16
  - U32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 22
stars: 0
tags:
  - mlx
  - inkling
  - moe
  - text-generation
  - speech
  - vision
  - text
  - audio
---

# Inkling-NVFP4-mlx-4bit

> Inkling-NVFP4-mlx-4bit - mlx-community 在 ModelScope 开源的模型。Inkling-mlx (4-bit, text backbone)

mlx-community/Inkling-NVFP4-mlx-4bit 是 ModelScope 魔搭社区上的 176.5B 参数text-generation模型，采用 apache-2.0 许可，基于 thinkingmachines/Inkling-NVFP4 构建。

- **Repository**: mlx-community/Inkling-NVFP4-mlx-4bit
- **License**: apache-2.0
- **Tasks**: text-generation
- **Parameters**: 176.5B
- **Base model**: thinkingmachines/Inkling-NVFP4
- **Tags**: mlx, inkling, moe, text-generation, speech, vision, text, audio
- **Downloads**: 22
- **Stars**: 0
- **Last updated**: 2026-07-17

Source: https://www.modelscope.cn/models/mlx-community/Inkling-NVFP4-mlx-4bit

---

# Inkling-mlx (4-bit, text backbone)

An **MLX 4-bit** build of the encoder-free **text backbone** of Thinking Machines' **Inkling**
(975B-total / 41B-active MoE), for running natively on Apple Silicon with
[`mlx-lm`](https://github.com/ml-explore/mlx-lm).

This is created for people using a two Apple Mac Studio M3 Ultra with 192/512 GB. (for one Mac Studio you need 2-bit!)

> **Community note** This is **not fullly numerically-verified**
> conversion, shared to see whether anyone can load/run a model this large on Apple
> Silicon and to gather feedback. Expect rough edges; please open a discussion with
> results (or failures).

## Notes

- **Memory:** the build is ~**580 GB** on disk (4-bit routed experts + bf16 attention /
  shared experts / embeddings). Loading it needs roughly that much **unified memory**, beyond
  any single Mac today (max 512 GB), so realistically it needs distributed/multi-device MLX or
  a very large box. This is largely a **research artifact**.
- **Not fully verified yet:** the custom Inkling forward (factorized attention + short-conv + sigmoid
  MoE) is a from-reference reimplementation whose logits have **not** been checked against the
  original. Correctness is unconfirmed.


## Provenance

- **Source:** `thinkingmachines/Inkling-NVFP4` (NVFP4) → dequantized → MLX affine **4-bit**
  (group size 64). Only the routed MoE experts are quantized; everything else is bf16.
- NVFP4→INT4 re-quantization, so expect small extra quality loss vs. a BF16-sourced build.

## Usage (once a loader is available)

```python
from mlx_lm import load, generate
model, tokenizer = load("mlx-community/Inkling-NVFP4-mlx-4bit")
print(generate(model, tokenizer, prompt="The capital of France is", max_tokens=64))
```

> The custom model class lives in the conversion repo (`models/inkling_mlx.py`); until it's
> registered in `mlx-lm`, load via that module's `load()`.
>
> Blog: https://huckiyang.github.io/blog/inkling-audio-design.html