#AI Models — Choosing Your Expert

Not all AIs think alike. ambientChat gives you access to multiple AI providers — each with different strengths — and lets you switch between them mid-conversation. It's like having a panel of experts on call, except they never argue with each other.

#What It Does

ambientChat connects to three major AI providers, each offering their own models:

• OpenAI (GPT-4o) — Excellent at general knowledge, structured output, and following precise instructions. The reliable all-rounder.
• Google Gemini — Strong at multimodal tasks (understanding images alongside text), fast responses, and broad factual knowledge. Good when you're sending photos along with questions.
• Anthropic Claude — Known for nuanced reasoning, careful analysis, and thoughtful longer-form responses. The one you want when you need to think something through.

The key insight: all models get the same context. Your inventory, documents, location, conversation history — everything flows to whichever model you choose. Switching models doesn't mean starting over. It means getting a second opinion from a different perspective, with all the same information.

Think of it as asking three different smart friends the same question. They all know your situation equally well, but they might frame their answers differently.

#How to Use It

#iOS App

#Switching Models in Chat

1. Open the Chat tab
2. Look near the top of the screen — the current model name is displayed (e.g., "Gemini", "GPT-4o", "Claude")
3. Tap the model name to open the Model Selector
4. A modal appears showing available AI providers and their models
5. Tap the model you want to switch to
6. Your next message (and all subsequent messages in this conversation) will be handled by the new model

The switch is immediate. You don't lose context, you don't lose your conversation history, and you don't need to re-explain anything. The new model picks up right where the old one left off.

#Model Settings

1. Go to Account (tap your profile icon or the Account tab)
2. Find AI Settings
3. Here you can:
   • Set your default model — which AI handles your questions when you start a new conversation
   • Enable or disable providers — turn off providers you don't want to see in the selector
   • View model capabilities — see at a glance what each model is best at

#Via Chat

You don't need to tell the AI which model to use via chat — just use the model selector. But you can ask about models:

• "Which AI model am I talking to right now?"
• "What are you good at?" (different models will answer this differently, which is itself informative)

And since all models share context, you can do things like:

1. Ask Gemini to analyze a photo of your pantry
2. Switch to Claude and say: "Based on what was just identified, suggest a recipe using those ingredients"
3. Switch to GPT-4o and say: "Now write me a shopping list for anything I'm missing for that recipe"

Each model builds on the conversation naturally.

#Tips & Tricks

• Try different models for different types of questions. There's no single "best" model — it depends on what you're asking. Quick factual lookups? GPT-4o is snappy. Need to reason through whether to keep or toss expired items? Claude thinks it through carefully. Sending a photo of a cluttered shelf? Gemini handles images well.
• Some models respond faster than others. If speed matters more than depth (checking a quick fact vs. analyzing your entire electronics collection), try the faster model first. You can always follow up with a more thorough model if the answer feels thin.
• Your token usage varies by model. Different models consume tokens at different rates. More capable models generally cost more per interaction. If you're watching your usage, keep an eye on the usage indicator in chat — it updates after each response.

#Options

#Known Limitations

• Not all models are available on all subscription tiers. Free tier users have access to a subset of models. Premium models may require a Standard, Premium, or Max subscription. The model selector shows which models are available to you.
• Some models cost more tokens than others. A single question to a more capable model may use more of your monthly token budget than the same question to a lighter model. The usage indicator in chat helps you track this.
• Model capabilities change over time. AI providers regularly update their models. A model that was best at image analysis last month might be overtaken by a competitor's update next month. We update ambientChat to support new model versions as they become available.
• Context window sizes vary. Each model has a limit on how much context it can process at once. For very long conversations or users with extremely large inventories, some models may truncate older context. Starting a fresh conversation resets this.
• Responses may differ between models for the same question. This is a feature, not a bug. Different models genuinely reason differently. If you get a surprising answer, try asking another model the same question and compare.

#Version History