Cat Llama 3 70B Instruct
Immersive roleplay via strict prompt adherence, Llama 3 70B finetune
8K contextMac
Local Models for Roleplay & Creative Writing
Models tuned for character roleplay, persona chat, and long-form creative writing — running fully offline on your iPhone, iPad, or Mac inside Private LLM, with nothing leaving your device.
13 models
EVA LLaMA 3.33 70B v0.1
Roleplay and creative writing specialist based on Llama 3.3 70B
131K contextMac
Llama 3.3 70B Euryale v2.3
Roleplay-optimized Llama 3.3 70B with flexible persona handling
131K contextMac
EVA Qwen2.5 32B v0.2
Roleplay and storywriting finetune of Qwen 2.5 32B with flexible tone control
131K contextMac
EVA Qwen2.5 14B v0.2
Roleplay and storytelling specialist fine-tuned from Qwen 2.5 14B
131K contextiPhone & iPad · Mac
Mythomax L2 13B
Roleplay specialist merging Llama 2 understanding and writing models
4K contextMac
Gemma 2 Ifable 9B
Roleplay specialist ranked first for creative writing on Gemma 2 9B
8K contextiPhone & iPad · Mac
Hathor_Stable v0.2 L3 8B
Roleplay specialist fine-tuned from Llama 3 8B Instruct
8K contextiPhone & iPad · Mac
L3 Umbral Mind RP v3.0 8B
Roleplay model for trauma and self-harm narratives on Llama 3 8B
8K contextiPhone & iPad · Mac
Llama 3.1 8B Lexi Uncensored V2
Uncensored Llama 3.1 8B finetune with safety refusals removed
131K contextiPhone & iPad · Mac
EVA Qwen2.5 7B v0.1
Roleplay finetune of Qwen 2.5 7B for adaptive persona and storytelling
131K contextiPhone & iPad · Mac
Samantha 1.2 Mistral
Roleplay finetune of Mistral 7B for caring interpersonal dialogue
33K contextMac
EVA D Qwen2.5 1.5B v0.0
Roleplay-focused distillation of Qwen 2.5 with 131K context
131K contextiPhone & iPad · Mac
Frequently asked questions
Private LLM runs roleplay and character models on iPhone, iPad & Mac, fully on-device. Larger models like Cat Llama 3 70B Instruct, EVA LLaMA 3.33 70B v0.1, Llama 3.3 70B Euryale v2.3 need more memory, so the right pick depends on your device's RAM.
Yes. Every model on this page runs entirely on your Apple device inside Private LLM — no cloud, no logging, and nothing sent off-device.
Only to download a model once. After that it runs fully offline in Private LLM.