Instructions to use TheStageAI/Qwen3.5-2B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use TheStageAI/Qwen3.5-2B-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use TheStageAI/Qwen3.5-2B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TheStageAI/Qwen3.5-2B-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TheStageAI/Qwen3.5-2B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Ollama
How to use TheStageAI/Qwen3.5-2B-GGUF with Ollama:
```
ollama run hf.co/TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M
```

Unsloth Studio

How to use TheStageAI/Qwen3.5-2B-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for TheStageAI/Qwen3.5-2B-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for TheStageAI/Qwen3.5-2B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for TheStageAI/Qwen3.5-2B-GGUF to start chatting

How to use TheStageAI/Qwen3.5-2B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use TheStageAI/Qwen3.5-2B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

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 "TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

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

Docker Model Runner
How to use TheStageAI/Qwen3.5-2B-GGUF with Docker Model Runner:
```
docker model run hf.co/TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M
```

Lemonade

How to use TheStageAI/Qwen3.5-2B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.5-2B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use TheStageAI/Qwen3.5-2B-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

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 TheStageAI/Qwen3.5-2B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

Qwen3.5-2B-GGUF / README.md

michellemoorre

Update model cards and release artwork

e42cdd7 verified 19 days ago

preview code

Raw

History Blame

9.69 kB

	---
	license: apache-2.0
	base_model:
	- Qwen/Qwen3.5-2B
	base_model_relation: quantized
	library_name: llama.cpp
	pipeline_tag: text-generation
	thumbnail: https://huggingface.co/TheStageAI/Qwen3.5-2B-GGUF/resolve/main/assets/thestage-edge-models-header.png
	tags:
	- gguf
	- llama.cpp
	- quantized
	- mixed-precision
	- local-inference
	- qwen3.5
	---

	<p align="center">
	<img src="./assets/thestage-edge-models-header.png" width="100%" alt="TheStageAI Edge Models: the right model at every memory budget">
	</p>

	<h1 align="center">Qwen3.5 2B</h1>

	<p align="center">
	Four GGUF checkpoints for llama.cpp, from 738 MB to 2.01 GB.
	<br>
	<strong>Start with M: 1.07 GB and ≈100% of the BF16 instruction-strict IFEval score.</strong>
	</p>

	<p align="center">
	<a class="inline-block" href="https://github.com/TheStageAI/edge-lm"><img class="dark:hidden" src="./assets/cta-edge-lm-light.svg" width="146" height="42" alt="Explore edge-lm on GitHub"><img class="hidden dark:block" src="./assets/cta-edge-lm-dark.svg" width="146" height="42" alt="Explore edge-lm on GitHub"></a>
	<a class="inline-block" href="https://docs.thestage.ai/"><img class="dark:hidden" src="./assets/cta-docs-light.svg" width="120" height="42" alt="Read TheStageAI documentation"><img class="hidden dark:block" src="./assets/cta-docs-dark.svg" width="120" height="42" alt="Read TheStageAI documentation"></a>
	<a class="inline-block" href="https://app.thestage.ai/"><img class="dark:hidden" src="./assets/cta-platform-light.svg" width="146" height="42" alt="Open TheStageAI Platform"><img class="hidden dark:block" src="./assets/cta-platform-dark.svg" width="146" height="42" alt="Open TheStageAI Platform"></a>
	</p>

	## Choose a checkpoint

	\| Tier \| Size \| Best for \| File \|
	\| --- \| ---: \| --- \| --- \|
	\| XS \| 738 MB \| Minimum footprint \| [Download](./Qwen3.5-2B-XS-TS-Q3_K_S.gguf) \|
	\| S \| 967 MB \| Compact \| [Download](./Qwen3.5-2B-S-TS-Q4_K_S.gguf) \|
	\| M \| 1.07 GB \| Recommended \| [Download](./Qwen3.5-2B-M-TS-Q4_K_M.gguf) \|
	\| L \| 2.01 GB \| High-precision Q8 \| [Download](./Qwen3.5-2B-L-TS-Q8_0.gguf) \|

	Other Qwen 3.5 sizes: <a href="https://huggingface.co/TheStageAI/Qwen3.5-0.8B-GGUF">0.8B</a>  ·  <a href="https://huggingface.co/TheStageAI/Qwen3.5-4B-GGUF">4B</a>  ·  <a href="https://huggingface.co/TheStageAI/Qwen3.5-9B-GGUF">9B</a>

	## Quickstart

	```bash
	llama-cli \
	--hf-repo TheStageAI/Qwen3.5-2B-GGUF \
	--hf-file Qwen3.5-2B-M-TS-Q4_K_M.gguf
	```

	## Why we recommend M

	At 1.07 GB, M matches the BF16 instruction-strict IFEval score within evaluation variation. It uses 47% less disk than L.

	\| Tier \| IFEval P / I (%) \| MMLU-Pro (%) \|
	\| --- \| ---: \| ---: \|
	\| BF16 reference \| 65.43 / 74.70 \| — \|
	\| L \| 65.80 / 74.94 \| — \|
	\| M \| 66.54 / 75.54 \| — \|
	\| S \| 63.22 / 73.38 \| — \|
	\| XS \| 52.68 / 64.15 \| — \|

	P / I means prompt-strict / instruction-strict. IFEval uses deterministic non-thinking decoding; MMLU-Pro uses sampled long-form reasoning. Only complete model-level scores are reported. A dash means not reported.

	> Reasoning: XS is intended for non-thinking use. Run it with `--reasoning off`. Choose S, M, or L for long-form reasoning.

	<details>
	<summary><b>Evaluation protocol</b></summary>

	- IFEval: 541 prompts, native chat template, `enable_thinking=false`, temperature 0.
	- MMLU-Pro: 12,032 questions, vLLM, 0-shot, native chat template `qwen_mc_json_v1`, `enable_thinking=true`; `temperature=1`, `top_p=0.95`, `top_k=20`, `min_p=0`, `presence_penalty=1.5`, `frequency_penalty=0`, `repetition_penalty=1`, `seed=42`; `max_model_len=40960`, `max_new_tokens=32768`; dataset revision `b189ec765aa7ed75c8acfea42df31fdae71f97be`.
	- The headline BF16 comparison uses instruction-strict IFEval; the raw scores are shown in the table.

	In matched long-form reasoning diagnostics, XS produced longer trajectories and reached the 32,768-token output limit more often than S. For Qwen 0.8B and 2B, a complete comparable MMLU-Pro matrix was not available at release time. The table omits incomplete subject runs.

	</details>

	## How the checkpoints are built

	All four checkpoints use the same production PTQ pipeline. The precision map is the only tier-specific part.

	### 1. Fit native GGUF codes

	Calibration activations produce a curvature objective weighted by true Fisher information for each quantized projection. We adapt the scale and minimum initialization from [NeUQI](https://arxiv.org/abs/2505.17595) to that objective, then solve integer codes on the target GGUF grid with a guarded cyclic coordinate-descent solver inspired by [QuantEase](https://arxiv.org/abs/2309.01885). A final K-quant pass tunes stored scales and minima while keeping packed codes fixed.

	### 2. Reconstruct the deployed trajectory

	Layers are calibrated in execution order against activations from the already-quantized prefix. A dense reference path measures accumulated drift. [Quantization Error Propagation (QEP)](https://arxiv.org/abs/2504.09629) adds that drift to the next reconstruction target, so later layers optimize for the inputs they receive at inference time.

	### 3. Allocate the byte budget

	XS and S can choose Q2_K through Q8_0 for each quantizable group. The optimizer trades changes in the teacher distribution against the actual encoded byte cost, including scale and minimum metadata. [ANNA](https://docs.thestage.ai/qlip/docs/source/anna_api.html) provides constrained configuration search. [RCO](https://arxiv.org/abs/2605.00649) provides an exact-budget route ([code](https://github.com/IST-DASLab/RCO)). For this model, ANNA selected both the XS and S precision maps. M and L keep the decoder qtypes at Q4_K_M and Q8_0, respectively, and use the same reconstruction and scale-tuning stages.

	After schedule selection, PTQ runs again from the source weights. Each layer is then calibrated with the final upstream precision choices.

	### 4. Align the full model

	A short affine distillation pass tunes native FP16 scales and minima while qtypes, packed codes, dense weights, and tensor layouts stay fixed. The loss matches the teacher's next-token distribution without changing file size or runtime layout.

	We load-test the shipping GGUF and evaluate it on a held-out set of 3,072 sequences with next-token KL. [`release-manifest.json`](./release-manifest.json) records its SHA-256, downstream evaluation IDs, and tensor metadata. Final recommendations use complete-model benchmarks.

	<details>
	<summary><b>File details</b></summary>

	\| Tier \| Hub selector \| GGUF file type \| Whole-file BPW \|
	\| --- \| --- \| --- \| ---: \|
	\| XS \| `Q3_K_S` \| `MOSTLY_Q2_K` \| 3.140 \|
	\| S \| `Q4_K_S` \| `MOSTLY_Q2_K` \| 4.109 \|
	\| M \| `Q4_K_M` \| `MOSTLY_Q4_K_M` \| 4.562 \|
	\| L \| `Q8_0` \| `MOSTLY_Q8_0` \| 8.558 \|

	The Hub selector controls sidebar grouping and download discovery. For XS and S it approximates the whole-file size class; [`release-manifest.json`](./release-manifest.json) contains the exact tensor mix. M and L keep their decoder qtypes at Q4_K_M and Q8_0 within the same production PTQ pipeline.

	Runtime memory also includes KV cache and buffers, which grow with context length.

	</details>

	## TheStageAI deployment stack

	These GGUF files target llama.cpp-compatible runtimes. [edge-lm](https://github.com/TheStageAI/edge-lm) runs compressed MLX models on Macs and iPhones. [ANNA](https://docs.thestage.ai/qlip/docs/source/anna_api.html) searches compression configurations under size or compute constraints. The [TheStageAI Platform](https://app.thestage.ai/) and [documentation](https://docs.thestage.ai/) cover compression, compilation, and serving workflows.

	For a specific device, latency target, or memory budget, [talk to our team](https://app.thestage.ai/contact).

	## Reproducibility

	- Release: July 21, 2026.
	- Base model: [`Qwen/Qwen3.5-2B`](https://huggingface.co/Qwen/Qwen3.5-2B) at revision [`15852e8c`](https://huggingface.co/Qwen/Qwen3.5-2B/tree/15852e8c16360a2fea060d615a32b45270f8a8fc).
	- Manifest: [`release-manifest.json`](./release-manifest.json) records the exact base revision, byte sizes, GGUF file types, whole-file BPW, tensor inventories, SHA-256 digests, held-out KL values, and evaluation IDs.
	- Runtime gate: export and load checks used [llama.cpp revision `bec4772f`](https://github.com/ggml-org/llama.cpp/commit/bec4772f6a2527d371557b5d2032641e5ff7619c).

	## Citation

	If you use this checkpoint, cite this release and follow the upstream model's citation guidance:

	```bibtex
	@misc{thestageai2026qwen3p52bgguf,
	author = {{TheStageAI}},
	title = {Qwen3.5 2B: TheStageAI GGUF Release},
	year = {2026},
	month = {jul},
	howpublished = {Hugging Face model release},
	url = {https://huggingface.co/TheStageAI/Qwen3.5-2B-GGUF},
	note = {XS, S, M, and L deployment tiers},
	}
	```

	<details>
	<summary><b>References</b></summary>

	- [ANNA](https://docs.thestage.ai/qlip/docs/source/anna_api.html), TheStageAI's constrained compression configuration search.
	- [RCO: Model Compression with Exact Budget Constraints via Riemannian Manifolds](https://arxiv.org/abs/2605.00649) ([code](https://github.com/IST-DASLab/RCO)).
	- [NeUQI: Near-Optimal Uniform Quantization Parameter Initialization for Low-Bit LLMs](https://arxiv.org/abs/2505.17595).
	- [QuantEase: Optimization-based Quantization for Language Models](https://arxiv.org/abs/2309.01885).
	- [Quantization Error Propagation: Revisiting Layer-Wise Post-Training Quantization](https://arxiv.org/abs/2504.09629).

	</details>

	## License

	The checkpoint weights use the upstream model's Apache-2.0 license. llama.cpp and other runtime software keep their own licenses.