Instructions to use AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit with MLX:

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

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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 "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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": "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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 "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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 "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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"

MLX LM

How to use AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Hermes Agent

How to use AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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 "AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-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 AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit

Run Hermes

hermes

Atomic Chat

AX-Ornith-1.0-35B-MLX-AXQ-6bit

Development AXQuant (AXQ) 6bit MLX pack of deepreinforce-ai/Ornith-1.0-35B @ 5df2ed3f675c7beaa490328cc70bb573b65fb660.

Claims (read carefully)

Claim	Status
AXQuant architecture-prior / development quant	Yes
Qwen 3.6 Tier 1 / Tier 2 certification	No
Measured coding-bench scores	Not claimed
MTP acceleration	Not claimed (no `-MTP` product label; source has no MTP weights)
Vision / VLM quality	Not claimed — vision tensors preserved at BF16 only

Source

Model: deepreinforce-ai/Ornith-1.0-35B (MIT; twin of ornith-ai/Ornith-1.0-35B)
Revision: 5df2ed3f675c7beaa490328cc70bb573b65fb660
Architecture: qwen3_5_moe 35B-A3B-class MoE
AXQuant adapter: qwen35-moe-v1 (convertible, development)
Prep: per-expert HF weights restacked to MLX packed experts.gate_up_proj / experts.down_proj before convert

Quantization

Toolkit: AXQuant (MLX-native PTQ)
Product class: 6bit
Planning: architecture-prior ladder (not measured sensitivity)
Optimization scope: text path only
Vision tower: BF16-preserved (vision.safetensors)

Intended use

Local Apple Silicon inference via MLX-LM for development evaluation. Do not treat this pack as a certified AutomatosX flagship release.

Attribution

Base weights © Ornith / DeepReinforce contributors under MIT. Quantization artifacts produced with AXQuant for research and development use.

Model tree for AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit

Base model

ornith-ai/Ornith-1.0-35B

Quantized

(182)

this model

Collections including AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit

AutomatosX MLX Model Catalog

Collection

Complete index of every Hub pack. Prefer the family collections above. • 93 items • Updated about 8 hours ago

Ornith 1.0

Collection

Ornith 1.0 35B AXQ 4/6-bit coding packs. • 2 items • Updated about 8 hours ago

AutomatosX
/

AX-Ornith-1.0-35B-MLX-AXQ-6bit

AX-Ornith-1.0-35B-MLX-AXQ-6bit

Claims (read carefully)

Source

Quantization

Intended use

Attribution

Links

Model tree for AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit

Collections including AutomatosX/AX-Ornith-1.0-35B-MLX-AXQ-6bit

AutomatosX MLX Model Catalog

Ornith 1.0