Skip to content
DeepSeek logo

Run Fuse O1 DeepSeek R1 QwQ SkyT1 32B on Mac

Reasoning model merging three chain-of-thought experts for math, coding, and science

Fuse O1 DeepSeek R1 QwQ SkyT1 32B runs 100% private on Mac inside Private LLM — no internet connection required, no data sent to any server.

32.8BMacreasoning

Model on Hugging Face

Specifications

Parameters
32.8B
Context window
131K tokens
License
Apache 2.0
Quantization
GPTQ-Int4
Family
DeepSeek R1 Distill

What Fuse O1 DeepSeek R1 QwQ SkyT1 32B is good at

FuseO1 DeepSeekR1 QwQ SkyT1 32B solves math, coding, and science problems by writing out a detailed chain-of-thought before giving an answer. It is a 32.8-billion-parameter model created by merging three existing open-source reasoning models (DeepSeek R1 Distill Qwen 32B, QwQ 32B Preview, and SkyT1 32B Preview) using a model fusion technique. The merged model accepts up to 131072 tokens of context. On the AIME24 math benchmark, it reached 74.0% pass@1 accuracy, outperforming the original source models and coming close to OpenAI o1. It also scored 62.1% on the GPQA Diamond science benchmark and 57.9% on the LiveCodeBench coding benchmark.

Which of your devices can run it

Mac

Mac (Apple Silicon, 192GB)MacBook Pro (M4 Max, 128GB)Mac Studio / Pro (Apple Silicon, 96GB)MacBook Pro (M4 Max, 64GB)MacBook Pro (M4 Max, 48GB)MacBook Pro (M4 Max, 36GB)Mac (Apple Silicon, 32GB)

Browse every model that fits iPhone or Mac.

How to run Fuse O1 DeepSeek R1 QwQ SkyT1 32B in Private LLM

  1. Download Private LLM from the App Store.
  2. Open the in-app model library and choose Fuse O1 DeepSeek R1 QwQ SkyT1 32B.
  3. Download the model once, then chat fully offline.
Download Private LLM on the App StoreJoin our Discord

Variants & related models

DeepSeek logo

DeepSeek R1 Distill Llama 70B

70.6B

Reasoning model distilling DeepSeek R1 reasoning into Llama 3.3 70B

131K contextMac
DeepSeek logo

DeepSeek R1 Distill Llama 8B

8B

Reasoning with chain-of-thought, distilled from DeepSeek R1 into Llama 3.1 8B

131K contextiPhone & iPad · Mac
DeepSeek logo

DeepSeek R1 Distill Llama 8B Abliterated

8B

Uncensored DeepSeek R1 Distill 8B via refusal abliteration

131K contextiPhone & iPad · Mac
DeepSeek logo

DeepSeek R1 Distill Qwen 14B

14.8B

Step-by-step reasoning distilled from DeepSeek R1 into Qwen2.5 14B

131K contextiPhone & iPad · Mac
DeepSeek logo

DeepSeek R1 Distill Qwen 32B Abliterated

32.8B

Uncensored proof-of-concept removing alignment guardrails from DeepSeek R1 Distill

131K contextMac
DeepSeek logo

DeepSeek R1 Distill Qwen 7B

7.6B

Step-by-step reasoning distilled from DeepSeek R1 into Qwen 7B

131K contextiPhone & iPad · Mac

Frequently asked questions

  • Fuse O1 DeepSeek R1 QwQ SkyT1 32B is too large for current iPhones. It runs on Mac with enough unified memory inside Private LLM, fully offline.

  • Yes. Fuse O1 DeepSeek R1 QwQ SkyT1 32B runs on Macs with enough unified memory, such as Mac (Apple Silicon, 192GB), MacBook Pro (M4 Max, 128GB), Mac Studio / Pro (Apple Silicon, 96GB), MacBook Pro (M4 Max, 64GB), MacBook Pro (M4 Max, 48GB), MacBook Pro (M4 Max, 36GB), Mac (Apple Silicon, 32GB), fully on-device in Private LLM.

  • Yes. Once downloaded in Private LLM, Fuse O1 DeepSeek R1 QwQ SkyT1 32B runs 100% on-device — no internet connection, and nothing is sent to any server.

  • Private LLM is a one-time purchase with no subscription and no per-message cost. The models themselves are open source — once downloaded, they run offline with nothing to pay per use.

Why run Fuse O1 DeepSeek R1 QwQ SkyT1 32B in Private LLM

Private LLM has run local AI on iPhone, iPad, and Mac since 2023, well before Apple Intelligence, LM Studio, Ollama, etc. existed. Inference happens on your device, so your Fuse O1 DeepSeek R1 QwQ SkyT1 32B conversations never reach a server. The part most apps gloss over is quantization, and that is exactly where on-device quality is won or lost. Most llama.cpp and MLX wrappers ship the same off-the-shelf 4-bit RTN weights. Private LLM ships GPTQ and OmniQuant quantization, tuned per model, and our 3-bit OmniQuant models match or beat those 4-bit RTN builds on the same Apple Silicon. Run the same model both ways and you feel it in the first reply. See how our quantization works.

Specifications and summary come from Fuse O1 DeepSeek R1 QwQ SkyT1 32B's Hugging Face model card, released under the Apache 2.0 license. Private LLM ships its own quantized models, built with GPTQ-Int4 quantization tuned per model, and isn't affiliated with the model's authors.