Choosing the Right AI Model ProThis feature requires the Pro plan
EquipDash gives admins eight AI models to choose from. They all do the same kinds of work — answering questions, looking up bookings, generating descriptions, helping with pricing — but they differ in speed, cost (how many credits each interaction uses), and how strong they are at complex reasoning.
This guide explains what each model is good at so you can pick one that fits your business.
Claude Sonnet 4.6 is our recommended default and works great out of the box. You can change it any time from Settings > AI Assistant > Configuration.
Quick comparison
| Model | Provider | Speed | Credit Cost | Needs Your Own Key? | Best For |
|---|---|---|---|---|---|
| GPT-5.4 Mini | OpenAI | Fast | Low | No | Quick everyday answers |
| GPT-5 | OpenAI | Slower | High | No | Complex reasoning |
| Claude Haiku 4.5 | Anthropic | Fast | Low | No | Quick tool calls |
| Claude Sonnet 4.6 (Recommended) | Anthropic | Slower | High | No | Best all-rounder |
| Kimi K2 | Moonshot | Fast | Lowest | No | Cheapest frontier model |
| Grok 4 Fast | xAI | Fast | — | Yes | Real-time information |
| Grok 4 | xAI | Slower | — | Yes | Premium real-time reasoning |
| Gemini 2.5 Flash | Fast | — | Yes | Fast, decisive on tool calls |
Recommendations by use case
Most businesses → Claude Sonnet 4.6 (default)
It's the strongest all-rounder. Reasoning is excellent, it follows instructions carefully, and it handles every kind of EquipDash task — from "show my bookings this week" to "draft a cancellation email for the customer waiting at pickup."
You don't need to do anything to use it — it's the default.
Fast and cheap → Claude Haiku 4.5 or GPT-5.4 Mini
If most of your team's questions are simple lookups ("how many bookings today?", "what's the email for John Smith?"), a faster, cheaper model gives the same answer for fewer credits. You'll get more usage out of your monthly allowance.
Lowest cost → Kimi K2
The cheapest frontier model in the picker. Surprisingly capable for tool-call work like creating bookings, looking up customers, or making recommendations. Good fit for high-volume teams.
Hardest reasoning tasks → GPT-5
When you have a really complex question — like "given my last 6 months of bookings, suggest a pricing strategy for the slow season" — GPT-5 does the most thorough analysis. Uses about 2× more credits than Sonnet, so save it for the hard stuff.
Real-time information → Grok 4 Fast or Grok 4 (your own key needed)
Grok models can pull current information from the web inside an answer (current weather, today's news, real-time prices). Useful if you ask the AI to compare your pricing against current market rates, or check weather before a booking.
You'll need a free xAI account to use these — see Use Your Own AI Account.
Multimodal & long context → Gemini 2.5 Flash (your own key needed)
Strong at handling lots of context (very long documents, image descriptions). Fast and decisive on tool calls. Available with a free Google AI Studio key.
How to switch models
- Go to Settings > AI Assistant.
- Click the Configuration tab.
- Under AI Model, click the model you want to use.
- Click Update to save.
The change applies to all AI features instantly — chat, generated descriptions, smart suggestions, agents, everything.
The picker only displays models you can actually use. If you haven't connected your own xAI key, Grok 4 Fast and Grok 4 won't appear. Same for Gemini. Once you add a key for that provider, the model becomes selectable.
Tips for choosing
- Start with the default. Claude Sonnet 4.6 covers nearly every use case well.
- Switch when you have a reason. "I want cheaper credits" → try Kimi or Haiku. "I want web data" → try Grok. "I want max reasoning" → try GPT-5.
- Different team members can't pick different models. The model is set at the supplier level — everyone on your team uses the same one.
- You can change it any time. No commitment. Try one for a day and switch back if you don't like it.