TalkToMe! (LLM Launcher)

TalkToMe! (LLM Launcher) combines on-device AI models, interactive 3D companions, and a pose and facial animation editor. Use it for conversations, writing, learning, creative work, or experimenting with local AI and character animation in one app.

ON-DEVICE AI MODELS
Download a model from the catalog or add a compatible GGUF model from your device or a download URL. Available text models include Qwen 3.5, LFM 2.5, and Gemma 4 variants, with different sizes and quantization levels to suit different hardware.

Use text models to explore ideas, explain concepts, draft and revise text, translate, role-play, or discuss code. Capabilities, supported languages, and response quality depend on the model you choose. Keep separate conversations with their own histories, switch chats, and manage downloaded models.

Compatible downloaded conversation models run inference locally on your device. You do not need a remote conversation API for local inference. Downloads, purchases, and online features still require an internet connection.

3D COMPANIONS AND EMOTIONS
Meet Fubuki, the built-in 3D companion, or import a compatible character model. Configure character identity and, with the relevant plan, personality instructions. Map compatible body bones and facial expressions to connect your character to the viewer's animation system.

AI replies can trigger emotional reactions such as happiness, sadness, surprise, or thinking. Bind a custom animation to an emotion to change how the companion responds visually. Reactions depend on the model's output and the character's available mappings; they are not real feelings or a guarantee of understanding.

POSE AND FACIAL ANIMATION EDITOR
Create animations with a fixed timeline, editable keyframes, and playback controls. Select a joint and rotate it with the 3D gizmo, adjust facial expressions, and combine body and face changes in the same sequence.

Save poses and expressions at specific frames, adjust transition timing, and preview with Play, Pause, and Stop. Manage, edit, or delete your custom animations. Creators Pack also lets you edit a custom copy of a discovered built-in animation while preserving the original.

APPEARANCE AND INTERACTION
Customize Fubuki with available outfits, hair and color options, accessories, backgrounds, lighting, and visual effects. Explore the character with the free camera, try gestures and touch interactions, and discover hidden animations. Earn eligible items through progression or access included content with the appropriate purchase. Fubuki-specific cosmetics do not transfer to arbitrary imported models.

VOICE AND IMAGE GENERATION
Use voice mode for hold-to-talk or continuous conversations, with compatible speech recognition and voice resources. Additional voices are available with the relevant plan. Language support and offline speech availability depend on your device and downloaded resources.

Generate images locally with a compatible image model, including the Stable Diffusion option in the catalog. Image generation uses a separate engine from text conversations and has its own loading and memory requirements.

HARDWARE AND ACCESS
Model files can be large. Available RAM, storage, processor, model architecture, quantization, and generation settings affect compatibility and speed. Not every GGUF model or 3D rig is supported. AI responses can be inaccurate; verify important information.

Some features, extra slots, voices, and content require in-app purchases or progression. Check the in-app plan details before purchasing. Imported characters need compatible rigs and facial mappings for animation features.