Documentation

Everything you need to get the most out of Perpendo.

Getting Started

Create your account at go.perpendo.app (free, no credit card).
Set an encryption passphrase in Settings — this protects your API keys.
Add at least one API key:
Start chatting.

Key Features

Model Switching

Click the model selector in the input bar to change models mid-conversation. Your conversation history stays intact.

Command Palette (Cmd+K)

Search conversations, prompts, agents, and actions from one place. Fuzzy search across everything you’ve ever discussed.

Conversation Branching

Hover a message and click the fork icon to branch. Explore different directions from any point in the conversation. Navigate branches with a visual tree.

Side-by-Side Comparison (Cmd+Shift+C)

Compare your chosen models with your own API keys. Send one prompt to 2–4 models, watch them stream in parallel, and branch from any response. Results are saved and private.

AI Agents

Save system prompt + model + parameter combos as one-click personas. Access from the sidebar Agents tab or via Cmd+K.

Prompt Library (Cmd+P)

Create, organize, and reuse prompts with {{variables}} that present input forms when triggered. Share across conversations.

Prompt Chains

Multi-step workflows where each step’s output feeds the next. Build in the Prompt Library → Chains tab. Example: Research → Outline → Draft in one click.

Web Search

Toggle the globe icon in the input bar. Perpendo searches the web, injects results as context, and the AI cites its sources. Requires a Tavily API key.

Voice Input

Click the mic icon or use on mobile to dictate messages.

Canvas Editor

Click “Edit” on any code block to open it in the side panel editor.

Analytics (Cmd+K → Analytics)

Track token usage, costs, and response times across all models. Export as CSV.

Tags & Archive

Add tags to conversations in the right panel and filter in the sidebar. Right-click a conversation → Archive (recoverable).

Export (Cmd+Shift+E)

Export conversations as Markdown or JSON.

Themes

Settings → Appearance to change themes or create custom ones.

Keyboard Shortcuts

Shortcut	Action
`Cmd+K`	Command palette
`Cmd+N`	New conversation
`Cmd+P`	Prompt library
`Cmd+,`	Settings
`Cmd+Shift+S`	Toggle sidebar
`Cmd+Shift+D`	Toggle right panel
`Cmd+Shift+C`	Toggle comparison mode
`Cmd+Shift+E`	Export conversation
`Cmd+/`	Show all shortcuts
`Cmd+.`	Stop generation
`Cmd+1-9`	Switch conversation
`Cmd+Enter`	Send message
`Shift+Enter`	New line
`Escape`	Close panel/modal

Local Models via Ollama (Self-Hosted)

Local model support requires the self-hosted version of Perpendo ($49 one-time license). The cloud version at go.perpendo.app uses remote API providers and cannot reach a localhost endpoint.

With the self-hosted version:

Install Ollama: brew install ollama
Start Ollama: ollama serve
Pull a model: ollama run llama3.1
In Perpendo, go to Settings → Custom provider → enter http://localhost:11434/v1
Your local models appear automatically.

Recommended models: Llama 3.1 8B (fast), Gemma 2 27B (powerful), Phi-4 14B (balanced).

Telegram Bot (Pro Plan)

Link your account in Settings → Integrations → Telegram.
Generate a link code, send /link CODE to the bot.
Chat with your AI models via Telegram.

Commands: /model, /agents, /new, /history. Send voice messages for transcription.

Plans & Billing

Free: 2 providers, 10 conversations, basic chat.
Plus ($5/mo or $39/yr): all features, all providers, unlimited conversations.
Pro ($9/mo or $79/yr): Plus + Telegram bot, prompt chains, web search.

Manage billing in Settings → Plan → Manage Billing.

FAQ

How do I change my passphrase?

Settings → Providers → update passphrase.

How do I delete my account?

Settings → Data → Delete Account.

Why can’t I add more providers?

The free plan is limited to 2 providers. Upgrade to Plus for all providers.

My messages aren’t streaming?

Check that your API key is valid and the provider is online. You can verify your key in Settings → Providers.

How do I import from ChatGPT?

Settings → Data → Import.