Instructions to use mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit")
config = load_config("mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit

Run Hermes

hermes

OpenClaw new

How to use mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "mlx-community/Devstral-Small-2-24B-Instruct-2512-6bit" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Fix tokenizer.json decoder: use ByteLevel so text decodes correctly

by DavidObando - opened 12 days ago

base: refs/heads/main

←

from: refs/pr/2

Discussion Files changed

-2

DavidObando

12 days ago

Hugging Face CLI

This repo's tokenizer.json uses a ByteLevel pre-tokenizer but its decoder is a Sequence (Replace ▁ / ByteFallback / Fuse / Strip) with no ByteLevel step, so byte-encoded tokens are never mapped back. As a result, decoded output contains literal Ġ (space) and Ċ (newline) artifacts, e.g. generation returns:

```pythonĊimportĠreĊ...

This also causes mlx_lm to select SPMStreamingDetokenizer, compounding the issue.

Fix: set the top-level decoder to ByteLevel ({"type":"ByteLevel","add_prefix_space":true,"trim_offsets":true,"use_regex":true}) — identical to the original mistralai/Devstral-Small-2-24B-Instruct-2512 and the working mlx-community/Devstral-Small-2-24B-Instruct-2512-4bit repos. Only the decoder field changes; vocab/merges are untouched.

Verified round-trip after the fix (spaces, newlines, indentation, unicode all decode correctly), and mlx_lm then selects BPEStreamingDetokenizer and produces clean output.

Fix tokenizer.json decoder: use ByteLevel so text decodes correctly68bfcad5

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

Cannot merge

This branch has merge conflicts in the following files:

tokenizer.json

· Sign up or log in to comment