AI providers
Updated
Keyword ideas, local phrases, listing drafts and the assistant run on an AI provider you choose. ASOMaster sells no AI credits: it uses a subscription you already pay for, or a model on your own computer.
Pick a provider
| Provider | What it uses | Install | Sign in |
|---|---|---|---|
| Claude Code | Your Claude Pro, Max, Team or Enterprise plan. No API key needed. | curl -fsSL https://claude.ai/install.sh | bash | claude auth login |
| Codex | Your ChatGPT Plus, Pro, Business or Enterprise plan. No API key needed. | npm install -g @openai/codex or brew install --cask codex | codex login |
| OpenCode | The model provider you connect in OpenCode: your own subscription or key. | curl -fsSL https://opencode.ai/install | bash | opencode auth login |
| Ollama | A model running on this computer, free. Needs a recent machine with enough memory. | brew install ollama | Start it with ollama serve, then ollama pull qwen3:8b |
| Anthropic API | Pay-per-use billing on your Anthropic Console account. Used only when you pick it. | Nothing to install | Paste an API key |
On Windows the installs use PowerShell, npm or winget instead; the app shows the right commands for your system.
Setup in the app
Open Settings › AI (or the first step of the setup guide). ASOMaster finds every copy of each CLI on your computer, including the ones bundled inside the Claude and ChatGPT desktop apps, and shows whether it's installed and signed in.
For any missing step, click to run it. The commands open in one Terminal window, one after another, so you can sign in in the browser and come back.
Click Test. ASOMaster asks the model a real question and shows how long it took. A provider only counts as Ready after that test passes.
Choose the model and, where the provider supports it, the reasoning effort (low, medium or high). With several installed copies, ASOMaster uses the newest by default; you can pin a specific one.
How the app uses it
- Fallback. If the chosen provider stops working (signed out, out of usage), ASOMaster can switch to another provider that is ready (turn this on or off in Settings › AI).
- Parallel work. One shared pool runs AI requests for all apps; Settings sets how many run at once.
- Clear errors. When a run fails, the app says why, for example not signed in, out of usage, the model isn't in your plan, or the installed copy is too old for the model, and offers the fix.
- No surprise bills. Subscription CLIs run without your API keys in their environment, so they can't quietly switch to pay-per-use billing. If a CLI is signed in with an API key instead of a plan, ASOMaster stops and tells you, unless you allow API billing for that provider in Settings.
- Read-only. CLIs run headless in a temporary folder with tools turned off or a read-only sandbox. They answer in a fixed format; they don't touch your files.
What is sent
Only the text a task needs: for example your app's name, current listing text, the product brief you wrote and the keywords with their numbers. It goes from your computer to the provider you chose, under that provider's terms; it doesn't pass through our servers. With Ollama, everything stays on your computer.
Your provider's own usage limits and charges still apply.