Frequently Asked Questions

We built Lookinglass to fight subscription fatigue. Instead of paying $20–$200/month whether you use AI or not, you only pay for the messages you send. See our transparent cost breakdown.

Costs vary by model and message length, but here are some examples:

• Simple questions (gpt-4o-mini): $0.001-0.01 per message
• Complex reasoning (o3): $0.01-0.10 per message
• Premium reasoning (o3-pro): $0.05-0.50 for complex tasks

The app shows you the maximum cost before sending each message, so there are never surprises. Learn about the cost bar.

Wallet funds are non-refundable but never expire. You can use them anytime, with no monthly fees or account maintenance charges. Whether you top up $1 or $100, the balance stays available until you need it.

Yes! We're confident you'll love Lookinglass, so all new users who sign in with Apple receive $1 in welcome credit to explore the app and try different models risk-free. It's our way of letting you experience the transparency and control that sets Lookinglass apart—no payment required to get started.

You probably shouldn't—if free ChatGPT's limitations work for you. See ChatGPT's pricing tiers.

ChatGPT Free tier gives you:

• 10 messages every 5 hours with gpt-5
• 1 gpt-5 Thinking message per day
• 16K context window (about 12,000 words)
• Automatic downgrades: After limits, switches to "mini version" without notification
• No visibility: Can't see which model you're using or how much capacity remains

ChatGPT Plus ($20/month) gives you:

• 160 messages every 3 hours with gpt-5
• 3,000 gpt-5 Thinking messages per week
• 32K context window (about 24,000 words)
• Model selection: Can manually choose models
• No pro models: No access to gpt-5-pro, o3-pro, or o1-pro

ChatGPT Pro ($200/month) gives you:

• Unlimited messages (subject to abuse guardrails)
• 128K context window for Fast mode, 196K for Thinking mode
• Access to gpt-5-pro
• No o3-pro or o1-pro: Those remain Enterprise-only

Consider Lookinglass if:

• You're a light user: Why pay $20 or $200/month when you can pay $2-$5/month?
• You want complete model access: Free tier gives you no selection. Plus ($20/month) gives you some models. Pro ($200/month) still doesn't include everything. Lookinglass gives you all 21 models with no subscription required.
• You need all pro models: Access o3-pro, o1-pro, AND gpt-5-pro without $200/month subscriptions or enterprise contracts
• You need context control: Set from 500 tokens to 1M tokens—you decide what the AI remembers
• Quality consistency matters: Choose your model, get consistent results every time—no surprise downgrades
• You need transparency: See your balance, costs, model, and tokens before every message

Start with $1 free credit and compare the experience yourself.

Lookinglass provides access to 21 OpenAI models including:

• ChatGPT Models (14 total): gpt-5 series (including gpt-5-pro), gpt-4o series (including chatgpt-4o-latest), gpt-4.1 series, gpt-4, and gpt-3.5
• O-Series Reasoning (7 total): o1, o1-pro, o1-mini, o3, o3-pro, o3-mini, and o4-mini

You can switch between models mid-conversation based on your needs. See how to switch models.

Transparency & Control: ChatGPT decides which model handles your request and automatically truncates conversation history to manage costs. With Lookinglass, YOU choose the model, YOU control context window size (500 tokens to 1M), and YOU see exactly what you're paying for.

Predictable Quality: ChatGPT routes to different models behind the scenes, so quality varies. Lookinglass gives you the exact model you selected—every time.

Fair Pricing: Subscriptions charge the same whether you send 10 messages or 1,000. With Lookinglass, light months cost less, heavy months cost more. Your bill matches your actual usage. Compare ChatGPT's plans.

Lookinglass gives you pay-per-use access to all three pro models: o3-pro, o1-pro, and gpt-5-pro.

Why this matters:

• o3-pro and o1-pro are Enterprise-only at OpenAI (no public pricing, requires sales contact)
• gpt-5-pro requires ChatGPT Pro ($200/month subscription)
• Lookinglass is the only way to access all 3 on a pay-per-message basis—no subscriptions, no enterprise contracts

Use pro models for your most important tasks, pay only when you need them. Learn about reasoning models.

Smart savings saves you 50% on message costs if you're willing to wait a bit longer for responses. It's available on select models (gpt-5 series, o3, o4-mini) and perfect for non-urgent tasks where you want to optimize for cost over speed. Try the pricing calculator to see potential savings.

Yes. We take privacy seriously — read our full privacy policy:

• No conversation storage on our servers: We don't store your messages or AI responses
• Local-first storage: Your conversations stay on your device
• Optional iCloud sync: If enabled, data syncs to YOUR personal iCloud account—we can't access it
• Minimal metadata: Only billing and basic usage info is kept
• Apple Sign-In: Secure authentication without exposing personal data

No, Lookinglass never stores your conversations or uses them for training. All prompts and responses stay on your device and are sent directly to OpenAI for processing via their API.

However, OpenAI may use API data to improve their models unless exempted by policy. We do not currently have an enterprise-level data exemption, so while we don't retain or view your data, OpenAI might use it to train future models.

For full transparency, see OpenAI's API Data Usage Policy.

We collect minimal technical data to support billing, performance monitoring, and app functionality. We never store your prompts, responses, or conversation history — all of that stays on your device.

• Model used and timestamp of each request
• Token counts (input and output) for accurate billing
• Selected settings (creativity level, AI power level, context window, output limit)
• Query latency and delivery status (e.g., success, timeout)
• Basic account details (if you sign in with Apple)

This data helps power features like cost tracking, usage history, and error handling; but we never store or access the content of your conversations.

Permanent memory lets you tell the AI key information about yourself once, and it remembers across all conversations. For example: "I'm a Python developer working on iOS apps." This context gets included automatically, so you don't need to re-explain your background each time.

Your memory is stored locally on your device and is sent with each message when enabled — but it's never stored on our servers or visible to us. You're always in control of what it remembers and when it's used. Learn more about memory and advanced controls.

Lookinglass gives you granular control over every aspect of your AI conversations. Here's what you can customize:

Smart Context (Smart Memory):

• Toggle on/off: Enable or disable conversation history entirely
• When enabled: Set from 500 tokens (minimum) up to your model's maximum context
• Adjustable in 500-token increments: Fine-tune exactly how much conversation history the AI remembers

Permanent Memory:

• Toggle on/off: Enable or disable persistent context across all conversations
• Custom text: Set background information about yourself that the AI remembers in every chat

Response Settings:

• Maximum response length: Set output tokens from 100 (minimum) in 100-token increments to control how long responses can be
• Live response streaming: Toggle whether responses appear letter-by-letter as they're generated or all at once

More context and longer responses cost more, but provide better continuity and more detailed answers. All conversation settings are accessible from the settings menu. Learn about advanced settings.

Context window size determines how much conversation history and text the AI can work with at once. Here's what different sizes mean in practical terms:

ChatGPT Free (16K tokens):

• ~12,000 words — about 20-30 pages of text
• Real example: Only 13% of The Hobbit (95,000 words)—you'd barely make it past Bilbo leaving the Shire
• In practice: Plenty for regular conversations, but paste in some code or debug logs and you'll max out fast

ChatGPT Plus (32K tokens):

• ~24,000 words — about 40-60 pages
• Real example: About 25% of The Hobbit—you'd make it to Rivendell, but not much further

ChatGPT Pro (128K tokens):

• ~96,000 words — a short novel
• Real example: The entire Hobbit, or about half of Fellowship of the Ring (187,000 words)

Lookinglass with gpt-4.1 series (1M tokens):

• ~750,000 words — multiple full-length books
• Real example: The entire Lord of the Rings trilogy (480,000 words) plus The Hobbit (95,000 words), with 175,000 words to spare—enough room for another full novel
• In practice: Analyze multiple books, long documents, or entire codebases in a single conversation without the AI forgetting earlier parts

Why this matters: When you hit the context limit, the AI starts "forgetting" earlier parts of your conversation. With 16K tokens, paste in a couple code files and suddenly the AI has amnesia about what you discussed five minutes ago. With 1M tokens, you can load entire codebases, multiple books, or extensive documentation and the AI remembers all of it.

For technical details about the gpt-4.1 series and 1M token context windows, see OpenAI's technical documentation.

Spending protection alerts you before sending messages that exceed your chosen threshold. Enable spending alerts in settings, set your preferred threshold (like $0.01, $0.10, or any amount), and you'll get a warning before accidentally sending an expensive message. See how cost transparency works.

Yes! Lookinglass includes built-in voice-to-text, so you can speak your questions naturally instead of typing them. Learn how to use voice input.

Yes, Lookinglass is completely free to download. You only pay for the messages you actually send.

No account is required to preview the app — guest mode lets you explore the interface and settings without signing in. However, guest mode does not allow you to send messages to AI models.

To interact with models and use wallet features, you'll need to sign in with Apple. This enables:

• Sending messages to any of the 21 AI models
• Accessing your wallet and transaction history across devices
• Using permanent memory features
• Saving and managing conversation threads

Just $1! You can add amounts from $1 to $500. Many users start with $5 or $10 to try it out.

Yes! Lookinglass includes an interactive AI Power User Guide that walks you through the app's key features when you first launch it. The tutorial covers:

• Chat Interface: Learn how to use the main conversation view
• Accessing Threads: Organize and navigate between different conversations
• Choosing AI Models: Browse and select from recommended models based on your needs
• Voice-to-Text: Use the microphone button for hands-free input
• Sending Messages: Tap to send and see real-time cost estimates in the cost bar
• Model Types: Understand different model categories (nano, mini, pro) and how reasoning models use reasoning tokens
• Conversation Settings: Toggle smart context and permanent memory on/off, control memory size (tokens), set maximum response length (tokens), choose billing plan, adjust creativity, response detail level, and AI power level
• Help Icons: Discover the question mark icons throughout the app for contextual explanations

You can replay the tutorial anytime from Settings → Your AI Power User Guide if you need a refresher.

Lookinglass is available in 5 languages:

• English
• Spanish (Español)
• French (Français)
• Portuguese (Português)
• Italian (Italiano)

The app automatically detects your device language and displays in the appropriate localization if it's one of the supported languages. All UI elements, settings, and help text are fully translated — only your conversations with AI models use the language you choose to type or speak in.

Lookinglass works on iPhone and iPad running iOS 17 or later. It's designed exclusively for Apple devices.

No, Lookinglass requires an internet connection to communicate with AI models. However, your conversation history and settings are stored locally when possible.

Yes! You can export entire chat threads in multiple formats including plaintext, JSON, or markdown documents. You can also copy and paste individual messages from your conversations. Additionally, you can export your transaction history and usage data as CSV files anytime. Learn how to export conversations.

Lookinglass includes a comprehensive audit system that gives you complete transparency into every transaction. The audit feature tracks two types of activities:

Wallet Top-Ups: See when you added funds to your account

API Usage (Message Costs): Each message you send is logged with detailed information including:

• Date & Time: Exact timestamp of when the request was made
• Model Used: Which AI model processed your message
• Status: Success, error, or failure indicator
• Request & Audit IDs: Unique identifiers for tracking and support
• Token Usage: Input tokens, output tokens, and cached input tokens (when applicable)
• OpenAI Rates Snapshot: The exact per-token pricing rates at the time of your request
• Cost Breakdown: Subtotal for OpenAI charges, Lookinglass fee, and total amount charged
• Device Information: Device type and device ID for multi-device tracking

This level of detail ensures you always know exactly what you paid for, which device made the request, and can verify charges against OpenAI's published rates. You can export your entire audit history as a CSV file for your records or expense tracking. See our transparent pricing breakdown.