Lookinglass Lookinglass
  • Home
  • Pricing
  • Guide
  • FAQ
  • Privacy
  • Terms
  • Download

Using Lookinglass: Your Complete Guide

Everything you need to know about chat, threads, and daily usage

Master the essentials of using Lookinglass—from your first message to organizing conversations and exporting your chat history. This guide covers everything you need to know about the chat interface and thread management.

Getting Started

Sign In

Your First Launch

When you first open Lookinglass, you'll see two options:

  • Sign in with Apple - Recommended for full access to all features, wallet functionality, and cloud sync
  • Continue as Guest - Explore the interface and settings without signing in (you won't be able to send messages to AI models)
Quick tip:

After signing in or continuing as guest, you'll land in the main chat interface—ready to start your first conversation.

The Chat Interface

Screen Layout

Screen Layout

The chat interface is designed to feel familiar—like WhatsApp or iMessage—but with powerful AI controls built in.

Top Bar:
  • Top Left Button - Opens your thread list to switch between conversations
  • Top Right Button - Opens the model picker to switch between 27 AI models
Chat History:
  • Your messages and AI responses appear here, organized chronologically
  • Live Streaming - Watch responses appear letter-by-letter as the AI generates them (if streaming is enabled in conversation settings)
Input Bar

The Input Bar

Below the message history, you'll find the input bar with four key components:

Left: Plus Button
  • Tap to upload images (up to 5 images per message)
  • Take photos or choose from your photo library
  • Most models support images (except gpt-3.5-turbo, gpt-4, and o3-mini)
  • Token counts include accurate image calculations
Center: Text Input Field
  • Type your message or prompt
  • Supports multi-line input for longer messages
  • Automatically expands as you type
Right: Microphone Button
  • Tap to use voice-to-text for hands-free input
  • First use requires granting permission for microphone and speech recognition
  • Speech recognition happens on-device for privacy
Right: Send Button
  • Paper airplane icon - the typical send button you'd see in any messaging app
  • Only becomes active once you've typed text or attached images
  • Tap to send your message to the selected AI model
  • Becomes a grayed-out Stop button when a message is being sent—tap to cancel anytime
  • Message is sent with your current conversation settings
Cost Bar

The Cost Bar

Below the input bar (and beneath the send button), Lookinglass displays a transparent cost estimate for every message before you send it. This is one of our core differentiators—you always know the maximum you'll pay, though your actual cost is often lower.

Left Side: Token Breakdown
  • Input Tokens (Estimate) - Shows how many tokens will be sent to the AI
  • Output Tokens (Maximum) - Shows the maximum response length based on your response settings
Right Side: Cost Information
  • Maximum Cost in USD - Always shows worst-case scenario (actual cost often lower), updates in real-time as you type
  • Info Icon (ⓘ) - Tap for a detailed breakdown of estimated tokens and costs
  • Pricing Icon ($) - Tap to open the full pricing sheet
Pricing Sheet

The Pricing Sheet

View complete pricing details for all 27 AI models at any time. Access the pricing sheet from two locations:

  • Model Menu - Tap the model picker (top right), then tap the circle with dollar sign ($) icon
  • Cost Bar - Tap the circle with dollar sign ($) icon to the right of the info icon
What the Pricing Sheet Shows:
  • OpenAI Base Prices - The original per-token rates from OpenAI
  • Lookinglass Prices - What you pay when using models through our app
  • Billing Plan Toggle - Switch between Fast & Reliable (standard pricing) and Smart Savings (50% off on 5 select models)

This level of transparency ensures you always know the pricing structure and can choose the billing plan that fits your needs. Learn more about transparent pricing or see typical costs.

Cached Input Savings

Why Your Actual Cost Might Be Lower

The cost bar shows the maximum cost for your message, but your actual cost may be significantly lower thanks to cached input tokens.

What are cached input tokens?
  • OpenAI stores your recent conversation context on their servers (typically for about 5 minutes)
  • When you send a follow-up message quickly, they don't need to resend all the previous conversation
  • Instead, they charge only for new tokens plus a much lower rate (up to 10X cheaper!) for cached tokens
Best way to save:
  • Have active, extended conversations in a single thread
  • Reply quickly within the conversation (within 5 minutes when possible)
  • The longer your thread and faster your replies, the more you'll save
Important to know:
  • Not all models support caching (pro models, chatgpt-4o-latest, and gpt-3.5-turbo don't support it)
  • Caching is not guaranteed — it can expire at any time
  • You can see when cached tokens were used in the message audit information

Think of it as getting a discount for continuing a conversation rather than starting fresh. Learn more about cached input.

Web Search

Web Search and Token Costs

When using models that support web search, enabling this feature can significantly increase your input token count.

Token impact:
  • Web search typically adds 10,000-25,000 input tokens or more
  • The exact amount depends on your question and the model you're using
  • Some complex queries can add 50,000+ input tokens
Cost control:
  • You can enable or disable web search anytime in conversation settings
  • Disable web search when you don't need current information to save on input tokens
  • The cost bar shows the estimated token count before you send, so you can see the impact

Web search is perfect for questions requiring up-to-date information, but disabling it when not needed is an effective way to reduce costs. Learn which models support web search.

Switching Models

Switching Models Mid-Conversation

Unlike traditional AI chat apps, Lookinglass lets you switch between any of the 27 models at any time—even mid-conversation.

  • Tap the model picker in the top right
  • Browse and select a different model
  • Continue the conversation with the new model
  • The cost bar updates instantly to reflect the new model's pricing
Use case:

Start with a fast, cost-effective model like gpt-4o-mini, then switch to o3-pro when you need deep reasoning on a specific question. Learn which model to choose.

Message Actions

Interacting with Messages

Long-press any message in the chat history to access message actions. You can also cancel any message (streaming or non-streaming) by tapping the Stop button that appears when a message is being sent.

Copy
  • Copy the message text to your clipboard
  • Paste it anywhere you need
Share
  • Use iOS's native share sheet to send the message via Messages, Mail, Notes, etc.
Info
  • View detailed audit information for that specific message
  • See exact token counts, model used, cost breakdown, and timestamp
Forward
  • Select the message, then tap Forward
  • Multi-select additional messages you want to include
  • Choose to either Share or Create New Thread

Perfect for branching off conversations—pick relevant messages and continue in a new thread.

Managing Threads

Threads in Lookinglass work like conversations in your messaging app—each one is a separate chat with its own history, model selection, and settings.

Thread List

Accessing Your Threads

Tap the top left button in the chat interface to open your thread list. Here you'll see:

  • Pinned threads - Appear at the top for quick access
  • Recent threads - Sorted by most recently active
  • Search - Find threads by name or content
Create Thread

Creating a New Thread

To start a fresh conversation:

  1. Open the thread list (tap top left button in chat interface)
  2. Tap the plus (+) button in the top right of the thread view
  3. Start chatting with a clean slate—no history from previous conversations
Note:

Each new thread starts with your default model and settings, which you can customize at any time. Learn about conversation settings.

Organize Threads

Organizing Threads

Keep your conversations organized with these thread management features:

Rename a Thread
  • Long-press or swipe on a thread in the list
  • Select Rename
  • Give it a descriptive name (e.g., "Python Project Help" or "Travel Planning")
Pin Important Threads
  • Long-press or swipe on a thread
  • Select Pin
  • Pinned threads stay at the top of your list for instant access
Copy a Thread
  • Long-press or swipe on a thread
  • Select Copy
  • Duplicate the entire conversation to continue exploring different directions
Archive Threads
  • Archive single: Swipe or tap Archive on a thread
  • Archive multiple: Long press to select multiple threads, then tap Archive
  • Archive all: Tap the top-right button (ellipsis menu) and select Archive All
  • Archived threads are moved out of your main list but not deleted
  • Access archived threads from the Archive view
  • In Archive view: unarchive (single/multiple/all) or delete from archive
Delete Threads

Deleting Threads

Clean up your thread list by removing conversations you no longer need:

Delete a Single Thread
  • Long-press or swipe on a thread
  • Select Delete
  • Confirm deletion (this cannot be undone)
Delete All Threads
  • Open the thread list
  • Tap the ellipsis (⋯) button in the top right (to the right of the + icon)
  • Select Delete All from the dropdown menu
  • Confirm you want to delete all conversations
  • Warning: This permanently deletes all threads and cannot be undone

Note for Multi-Device Users: If you have iCloud sync enabled and use Lookinglass on multiple devices, the ellipsis menu also includes a Refresh option to manually sync threads across your devices. Learn more about iCloud sync and privacy.

Search Threads

Searching for Threads

Find specific conversations quickly using the search feature:

  • Open the thread list
  • Use the search bar at the top
  • Search by thread name or message content
  • Results update in real-time as you type

Looking for specific threads? Learn how to export your conversation history.

Exporting Conversations

Lookinglass makes it easy to export your conversations for backup, sharing, or analysis. You own your data—take it with you anytime.

Export Formats

Available Export Formats

Export entire threads in your preferred format:

Markdown
  • Clean, readable format perfect for documentation
  • Preserves formatting and structure
  • Easy to edit and share with others
  • Great for integrating into wikis, GitHub, or note-taking apps
Plaintext
  • Simple, universal format that works everywhere
  • No formatting—just pure conversation content
  • Smallest file size
  • Perfect for archiving or text analysis
JSON
  • Structured data format for developers
  • Includes metadata (timestamps, model used, token counts)
  • Machine-readable for automation or custom processing
  • Ideal for building your own analysis tools
How to Export

How to Export

To export a thread:

  1. Open the thread list (top left button)
  2. Long-press or swipe on the thread you want to export
  3. Select Export
  4. Choose your format: Markdown, Plaintext, or JSON
  5. Share or save the file using iOS's share sheet
Additional exports:

You can also export your entire transaction history and usage data as CSV files from the wallet section for expense tracking and record-keeping.

Need just a few messages?

Use the Forward feature to select specific messages and either share them or create a new thread with just that subset of your conversation.

Using Voice Input

Lookinglass includes built-in voice-to-text for natural, hands-free interaction with AI models.

Privacy & Setup

First-Time Setup

When you first tap the microphone button, iOS will prompt you to grant two permissions:

  • Microphone Access - Required to capture your voice
  • Speech Recognition - Required to convert speech to text
Privacy note:

These permissions are used only for voice input and all speech recognition happens on-device for your privacy. Lookinglass never records or stores your voice. Learn more about privacy.

Using Voice

Using Voice-to-Text

  1. Tap the microphone button inside the text input bar on the bottom right
  2. Speak your message clearly
  3. Your speech is transcribed in real-time to the text input field
  4. Tap the send button when ready, or edit the text first
Language Support:

Voice input uses your device's native speech recognition (provided by Apple), so it works in any language your device supports. The Lookinglass interface itself is available in 5 languages: English, Spanish, French, Portuguese, and Italian.

Tips

Tips for Best Results

  • Speak clearly - Enunciate for more accurate transcription
  • Minimize background noise - Find a quiet environment when possible
  • Review before sending - Check the transcribed text for accuracy
  • Edit as needed - You can tap the text field to manually correct any mistakes

Next Steps

Now that you understand the basics of using Lookinglass, explore these topics to get even more from the app:

Models

Model Guide

Learn about all 27 AI models and when to use each one

Settings

Conversation Settings

Control smart context, permanent memory, and response length

Protection

Spending Protection

Set alerts to avoid accidentally expensive messages

Pricing

Pricing

Understand how pay-as-you-go works and compare to subscriptions

Need more help? Contact us at [email protected]

Lookinglass Lookinglass

Product

  • AI Models
  • Features
  • Download

Support

  • Contact
  • FAQ

Legal

  • Privacy Policy
  • Terms of Service

© 2025 Lookinglass. All rights reserved.

See further, think deeper with Lookinglass