Choose a model, or let Auto choose
Knecty gives you models from several companies in one place. You can pick one yourself for every message, or let Auto pick.
The model picker
Open the picker from the model name in the message box. It is grouped by company — Auto first, then ChatGPT, Claude, Gemini, Grok, DeepSeek, Mistral and others — and each company shows its best-known models first, with a View all link for the rest.
Every row shows:
- the model name, and a ✨ if it can think before answering;
- a short note on what it’s good at;
- a typical price per message (for example “~$0.002/msg”).
Tap ⓘ on a row for the details: price per million tokens (the provider’s rate plus the 25% platform fee), how much text it can read at once (its context), speed, whether it reads images, and which company makes it.
Use Filters to sort by price or speed, or to show only models that read images (Vision), text models, or image generators. Your filter choice is remembered in this browser.
A row may be struck through:
- Out of credit — your balance can’t cover this model right now. Top up, or wait for your weekly credit.
- Unavailable — the provider isn’t serving it at the moment. Pick another model.
Switching mid-chat
You can change the model at any point in a conversation. The next reply comes from the new model, which reads the conversation so far. If the chat is too long for the model you picked, Knecty offers to compact it first — see Long chats and compaction.
Your default model
Settings → Models → Default model sets the model new chats start on. Choose a specific model, Auto (the best model per message across companies), or a company’s Auto, such as Auto · Claude. If you have never chosen, new chats start on Auto.
Models in your region
In the EU and UK, DeepSeek models are hidden and never used. See Models available in your region.