Instructions to use FINAL-Bench/POCKET-KR-MLX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FINAL-Bench/POCKET-KR-MLX 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("FINAL-Bench/POCKET-KR-MLX")

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 FINAL-Bench/POCKET-KR-MLX with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "FINAL-Bench/POCKET-KR-MLX"

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": "FINAL-Bench/POCKET-KR-MLX"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use FINAL-Bench/POCKET-KR-MLX 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 "FINAL-Bench/POCKET-KR-MLX"

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 FINAL-Bench/POCKET-KR-MLX

Run Hermes

hermes

OpenClaw new

How to use FINAL-Bench/POCKET-KR-MLX with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "FINAL-Bench/POCKET-KR-MLX"

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 "FINAL-Bench/POCKET-KR-MLX" \
  --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 FINAL-Bench/POCKET-KR-MLX with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "FINAL-Bench/POCKET-KR-MLX"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "FINAL-Bench/POCKET-KR-MLX"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "FINAL-Bench/POCKET-KR-MLX",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

POCKET-KR-MLX / README.md

SeaWolf-AI

POCKET family cross-links + Image/Studio/Zimage

a63101d verified about 8 hours ago

preview code

Raw

History Blame Contribute Delete

9.76 kB

metadata

license: apache-2.0
library_name: mlx
pipeline_tag: text-generation
base_model:
  - FINAL-Bench/Darwin-36B-Opus
tags:
  - conversational
  - on-device
  - mobile
  - iphone
  - android
  - cpu
  - local-llm
  - edge
  - mixture-of-experts
  - moe
  - quantized
  - pocket
  - vidraft
  - qwen3_5_moe
  - mlx
  - apple-silicon
  - korean
  - korean-llm
  - darwin

📚 Collections

▶ POCKET Models — this family (on-device, no GPU) Darwin Family · Aether Foundation · VKAE Accelerated · Metacognition Adapters

POCKET-KR-MLX · 🍎 iPhone / Mac

35B 한국어 모델을 아이폰에서 네이티브로. Apple MLX 2-bit, 5 GB. iPhone·iPad·Mac에서 MLX Swift로 바로 실행.

🚀 Try it live, no install → — both answering on a CPU-only box (no GPU). POCKET-26B is Gemma4-based.

Pick your build →

The POCKET lineup — pick by your device

Repo	File	Size	Runs on	Best for	Korean PPL*
POCKET-35B-GGUF	`Q4_K_M`	21 GB	PC / server (32 GB RAM)	top quality	5.79
POCKET-35B-GGUF	`Q2_K` ⭐	13 GB	mini-PC, no GPU	daily driver	6.49
POCKET-35B-GGUF	`IQ1_M`	8.2 GB	16 GB RAM box	smallest full model	9.69
POCKET-KR-GGUF	`IQ2_M`	5.1 GB	Android 8 GB+	🇰🇷 Korean phone	7.95
POCKET-KR-MLX	2-bit	5.1 GB	🍎 iPhone / iPad / Mac	🇰🇷 Korean, Apple-native	7.95
POCKET-EN-GGUF	`iPhone-mix`	5.3 GB	🍎 iPhone (PocketPal)	🌍 English phone	—
POCKET-EN-GGUF	`PC-mix`	6.8 GB	PC / Android	🌍 English, best quality	—

*Wikipedia-Korean perplexity, lower is better. Q4_K_M = 5.79 baseline. English builds are tuned on English; see each repo.

🍎 Why MLX for Korean but GGUF for English on iPhone? Apple-native MLX only does uniform quantization. Korean survives it (96 experts hold up); English needs our proprietary quantization, which only GGUF supports — so the English iPhone build ships as a GGUF you run with PocketPal. Honest, not lazy.

🆕 POCKET-26B — a Gemma4-26B-A4B-based sibling that loads in any app today (Ollama · LM Studio · PocketPal · MLX), no bleeding-edge runtime needed: GGUF (Q2_K 11 GB · Q4_K_M 17 GB · GPQA-Diamond 67%). Universal compatibility for 12 GB phones, PC, and browser.

Benchmarks — what is measured, what is not

We measure Bonsai on the same machine with the same stock llama.cpp, and we tell you where we lose.

[measured] Generation speed — POCKET wins on both CPU and GPU:

	POCKET-35B IQ1_M	Bonsai-27B Q1_0
CPU generate (Xeon, 16t)	27.0 tok/s	10.1	🟢 2.69×
GPU generate (H100)	197 tok/s	89	🟢 2.22×
GPU prompt (H100)	753	1816	🔴 0.41×
Quality (HellaSwag, 400q)	61.0%	60.0%	⚪ tie (CI overlaps)

[measured on a MacBook M3 Pro, 18 GB] — and on a laptop, POCKET wins every axis, including prompt processing:

	POCKET-35B IQ1_M	Bonsai-27B Q1_0
Metal generate (tg64)	25.4 tok/s	12.8	🟢 1.99×
CPU generate (8 threads)	13.8 tok/s	4.4	🟢 3.13×
Metal prompt (pp128)	240.7 tok/s	73.4	🟢 3.28×
CPU prompt (pp128)	45.5 tok/s	9.6	🟢 4.75×

On a laptop GPU the arithmetic headroom that let Bonsai win prefill on an H100 is gone, so MoE sparsity wins across the board. POCKET-35B-Q2_K runs on the M3 Pro's CPU at 19.5 tok/s — on an 18 GB Mac, run Q2_K on CPU (-ngl 0); its 13 GB exceeds the recommended Metal budget.

[measured — GPQA Diamond, 198q, greedy] reasoning quality vs quantization:

Model	GPQA-Diamond (greedy)
Qwen3.6-35B-A3B	73.2%
POCKET-35B Q4_K_M	68.7%
POCKET-35B Q2_K	60.1%

[pending — community reports welcome] on-device iPhone and Strix Halo throughput. We publish only what we ran ourselves; help us fill the rest.

The same-size rival Ternary-Bonsai-27B-Q2_0 (7.2 GB) fails to load in upstream llama.cpp — it needs the PrismML fork. POCKET runs on the tools you already have.

Files in this repo

Format	Size	Runs on
MLX 2-bit (`model-*.safetensors`)	5.1 GB	🍎 iPhone Pro / iPad / Mac

Apple-silicon native (Metal). For Android/PC use the GGUF build.

Quickstart (Mac)

pip install mlx-lm
mlx_lm.generate --model FINAL-Bench/POCKET-KR-MLX --prompt "안녕하세요"

On iPhone/iPad: MLX Swift examples.

⚠️ On-device speed is not yet measured by us — reports welcome.

Lineage — where POCKET comes from

POCKET is quantized from Darwin-36B-Opus, VIDRAFT's flagship — a model bred and evolved over several generations on the Darwin platform (crossbreeding, healing, expert surgery). Darwin-36B-Opus itself traces back to a Qwen3.5-family MoE architecture.

Component	Origin
Starting checkpoint	Darwin-36B-Opus — VIDRAFT, multi-generation Darwin evolution
Base architecture	Qwen3.5-family MoE (256 experts, top-8), unchanged
Quantization (`Q4_K_M`…`IQ1_M`)	stock llama.cpp — no custom format
Runtime	upstream llama.cpp / Apple MLX — unmodified
Proprietary language-specific tuning (KR/EN builds)	ours (VIDRAFT)

The CPU/GPU speed comes from the sparse-MoE architecture plus ordinary quantization — reproducible with the same base and the same tools. What we add is the Darwin-evolved weights, the honest measurement, the Korean tuning, and the pruning that makes the 5 GB phone builds.

Limitations

The iPhone/Mac speed is not yet measured by us — community reports welcome.
Extreme quants (IQ1_M) hurt Korean ~2.8× more than English; use Q2_K or larger for quality.
English phone builds trade quality for size; the PC build (PC-mix) is much closer to full quality.

License

Apache-2.0.

POCKET is a VIDRAFT model family. 35B, in your pocket. No GPU.

Learn more

On-device LLMs without a GPU — and how POCKET measures up: Can you run a large LLM without a GPU?
What model quantization is, and why a 4-bit model stays smart: What is model quantization?

🧩 The POCKET Family — On-device AI by VIDRAFT

Big models, small hardware. No GPU, no cloud.

Models

📦 POCKET-35B-GGUF — flagship, PC / server, no GPU
📦 POCKET-26B-GGUF — compact 26B
🇰🇷 POCKET-KR-GGUF — Korean, Android
🍎 POCKET-KR-MLX — Korean, iPhone / Mac
🌍 POCKET-EN-GGUF — English, phone / PC
🖼️ POCKET-Image-Zimage — character-perfect text in any image

Demos & tools (Spaces)

🎨 POCKET-Image Studio — text-in-image, generate in-page
🖥️ POCKET-35B-CPU — 35B answering on a CPU
🖥️ POCKET-26B-CPU — 26B on a CPU

📚 Full POCKET collection