Skip to content
Google Gemma logo

Run Gemma 2 9B IT SPPO Iter3 on iPad & Mac

General assistant refined from Gemma 2 9B via three rounds of self-play training

Gemma 2 9B IT SPPO Iter3 runs 100% private on iPad & Mac inside Private LLM — no internet connection required, no data sent to any server.

9.2BiPadMacgeneral

Specifications

Parameters
9.2B
Context window
8K tokens
License
Gemma
Quantization
OmniQuant 4-bit
Family
Gemma 2 9B

What Gemma 2 9B IT SPPO Iter3 is good at

Fine-tuned from the Gemma 2 9B instruction-tuned model, Gemma 2 9B IT SPPO Iter3 is a 9.2 billion parameter assistant that follows instructions and draws on general knowledge. It was optimized with Self-Play Preference Optimization over three training iterations, making it notably good at producing helpful responses aligned with human preferences.

Which of your devices can run it

iPad

iPad Pro (M5, 16GB)iPad Pro (M4, 16GB)iPad Pro (M2, 16GB)iPad Pro (M1, 16GB)

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)MacBook Air (M4, 24GB)MacBook Air (M-series, 16GB)

Browse every model that fits iPhone or Mac.

How to run Gemma 2 9B IT SPPO Iter3 in Private LLM

  1. Download Private LLM from the App Store.
  2. Open the in-app model library and choose Gemma 2 9B IT SPPO Iter3.
  3. Download the model once, then chat fully offline.
Download Private LLM on the App StoreJoin our Discord

Variants & related models

Google Gemma logo

FuseChat Gemma 2 9B Instruct

9.2B

General-purpose assistant distilled from four larger models into Gemma 2 9B

8K contextiPhone & iPad · Mac
Google Gemma logo

Gemma 2 9B IT

9.2B

General-purpose model in the Gemma 2 9B family

iPhone & iPad · Mac
Google Gemma logo

Gemma 2 Ifable 9B

9B

Roleplay specialist ranked first for creative writing on Gemma 2 9B

8K contextiPhone & iPad · Mac
Google Gemma logo

Tiger Gemma 9B v3

9.2B

Uncensored roleplay and text generation finetune of Gemma 2 9B

8K contextiPhone & iPad · Mac
Meta logo

Airoboros l2 7b 3.0

6.7B

Instruction following specialist built on Llama 2 7B

4K contextiPhone & iPad · Mac
Mistral logo

Airoboros M 7B

7.2B

General instruction-following model built on Mistral with structured math output

33K contextMac

Frequently asked questions

  • Gemma 2 9B IT SPPO Iter3 is too large for current iPhones. It runs on Mac with enough unified memory inside Private LLM, fully offline.

  • Yes. Gemma 2 9B IT SPPO Iter3 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), MacBook Air (M4, 24GB), MacBook Air (M-series, 16GB), fully on-device in Private LLM.

  • Yes. Once downloaded in Private LLM, Gemma 2 9B IT SPPO Iter3 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 Gemma 2 9B IT SPPO Iter3 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 Gemma 2 9B IT SPPO Iter3 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 Gemma 2 9B IT SPPO Iter3's Hugging Face model card, released under the Gemma license. Private LLM ships its own quantized models, built with OmniQuant quantization tuned per model, and isn't affiliated with the model's authors.