Chat Interface Guide
Everything you need to know about SOHAM's chat — from basic messaging to voice input, TTS, and advanced conversation techniques.
Core Features
Context-Aware Conversations
SOHAM remembers everything said in the current session. Ask follow-up questions naturally — no need to repeat context.
Smart Auto-Routing
Automatically selects the best AI model and capability for each message — web search, image generation, PDF analysis, or direct chat.
Voice Input
Speak your message using Groq Whisper V3 Turbo for fast, accurate speech-to-text transcription.
Text-to-Speech
Have responses read aloud using Groq Orpheus TTS with 6 natural-sounding voices.
Message Sharing & Export
Share individual messages or export the full conversation as text or markdown.
Live Web Search
DuckDuckGo search is auto-triggered for time-sensitive queries, or use /search explicitly.
Interface Overview
Message InputType your message here. Supports multi-line input with Shift+Enter.
Voice Button (🎤)Hold to record, release to transcribe. Powered by Groq Whisper V3 Turbo.
Model SelectorClick to switch between 13+ AI models mid-conversation.
Settings (⚙️)Adjust tone, technical level, TTS voice, and theme.
Clear ConversationWipe the current session from memory. Irreversible.
Export / ShareExport the conversation or share a specific message.
Voice Input Setup
Powered by Groq Whisper V3 Turbo (STT)
- 1
Allow microphone access
When prompted by your browser, click "Allow" to grant microphone permission.
- 2
Click the microphone button
The mic button is in the message input area. Click it to start recording.
- 3
Speak your message
Speak clearly. SOHAM uses Groq Whisper V3 Turbo for fast, accurate transcription.
- 4
Review and send
Your speech is transcribed into the input field. Edit if needed, then press Enter to send.
Whisper V3 Turbo supports multiple languages and handles accents well. Works best in a quiet environment.
Text-to-Speech Setup
Powered by Groq Orpheus TTS — 6 voices
- 1
Open Settings
Click the ⚙️ icon in the header to open the settings panel.
- 2
Enable Text-to-Speech
Toggle the TTS switch to on. The toggle turns blue when active.
- 3
Choose a voice
Select from 6 Orpheus voices: troy, diana, hannah, autumn, austin, or daniel.
- 4
Send a message
SOHAM will automatically read its next response aloud using your chosen voice.
Conversation Tips
Be specific
Instead of "explain sorting", try "explain merge sort vs quicksort for a job interview".
Use follow-ups
"Now show me a Python implementation" — SOHAM remembers the context from earlier in the chat.
Use slash commands
/solve for step-by-step working, /search for live web results, /summarize for condensed output.
Set your technical level
Set "Expert" in settings for dense, precise answers without hand-holding.
Advanced Features
Conversation Memory
Within-session context retention
SOHAM maintains full conversation history within a session. You can reference anything said earlier: "go back to the code you wrote in step 2" or "now apply that to the second example".
History is stored in your browser's memory and cleared when you close the tab or click "Clear conversation".
Personalised Responses
Settings that shape every reply
Your tone (Helpful / Formal / Casual) and technical level (Beginner / Intermediate / Expert) settings apply to every message in the session. Change them mid-conversation and the next response will reflect the new settings immediately.
Mobile Chat Experience
Touch-optimised for Android and iOS
Touch-optimised layout
Larger tap targets, swipe gestures, and a bottom-anchored input bar for thumb-friendly typing.
Bottom sheet model selector
Tap the model name to open a full-screen bottom sheet with all 13+ models listed.
Voice priority
The microphone button is prominently placed for quick voice input — ideal when typing is inconvenient.
PWA full-screen mode
Install as a PWA for a native app feel with no browser chrome. See the Installation guide.
Start chatting
No account needed. Open the chat and start typing — or speak your first message.