Local LLM Studio runs models on your own machine. These are the pieces it needs.
Pick a .gguf above and press Import into Ollama.
Waiting for output…
Use Gallery in the header (or Open Gallery) to browse Studio + local Forge outputs inside this app. Forge IIB opens the full extension in a new tab.
Large side avatars + chat wallpaper. Point each at a folder of images. Cycle 5–60s or static. Chat: /avatar me C:\pics · /avatar model … · /bg …
Defaults for new bubbles. Use Folder (cycle) or Image (static). Click any bubble avatar for per-bubble settings.
Applies to local and external chat. Also sets Max Tokens unless you change tokens after saving.
Real Google web + image results via
Serper.dev.
Free tier: sign up → API key. Use chat buttons or
/web … / /images ….
Local Kokoro-82M voices. Click Speak on any assistant reply, or enable auto-play.
Starts reading after the first sentence instead of waiting for the whole reply. Same as the 🔊 icon on a reply.
Welcome to Local LLM Studio! Select and load a model to start chatting.
New features:
/avatar me C:\pics · /avatar model … · /bg … set defaults/imagine … or the image button for Forge generationUses the Forge extension stable-diffusion-webui-wd14-tagger (local, needs --api). Drop an image, pick a gallery file, or tag the last chat image.
Images always open in a new bubble under the last reply (model text is kept).
Select a profile on the left, edit fields, then Save Changes. Your old prompts are in My Current Setup. Use Strict No-Refuse for the hard compliance profile.
Paste a JSON object (or contents of a .json / .txt file) below, then Apply.
Are you sure you want to delete this chat? This action cannot be undone.