// neonForge LIGHT MODE

dolphin-ui

unified AI interface -- chat / image gen / voice synthesis

deployed ec2:8200 voice loop: planning

overview

dolphin-ui is a three-service web interface that unifies self-hosted AI capabilities into a single dashboard. it connects a local LLM (Ollama on GPU), image generation (Stable Diffusion on T4), and text-to-speech (ElevenLabs API) behind a Flask backend on questboard-ec2.

services

chat -- Ollama / dolphin-llama3:8b

Streaming chat via Ollama REST API. Messages proxied from Flask to picass0.

model: dolphin-llama3:8b (Q4_K_M, ~6GB VRAM)
streaming SSE responses via /api/chat
conversation history maintained client-side

image gen -- Stable Diffusion v1.5 on T4

Image generation via custom Flask endpoint on picass0. Uses diffusers library with SD v1.5 in float16.

GPU: Tesla T4 (16GB VRAM)
resolution: 256-1024px
returns base64-encoded PNG

audio -- ElevenLabs TTS

Text-to-speech via ElevenLabs API. Voice list pulled dynamically. Audio returned as base64 MP3.

default voice: Rachel
API key from vault: elevenlabs_miecz_api_key
audio history with replay in the UI

API endpoints

method	path	description
POST	/api/chat	streaming chat -- accepts `{messages, model}`
GET	/api/models	list available Ollama models
POST	/api/image	image generation
POST	/api/tts	text-to-speech -- returns base64 audio
GET	/api/voices	list ElevenLabs voices
GET	/api/health	service health status

deployment

component	host	port
Flask app	questboard-ec2	8200
nginx proxy	questboard-ec2	8201
image gen API	picass0	7860
Ollama	picass0	11434

known issues

no git repo -- project is loose files
ElevenLabs keys may be expired
picass0 is on-demand -- requires GPU instance running
no auth on web UI

neonForge // dolphin-ui docs // goulard3120 // 2026-05-07 // LIGHT MODE