Open your project
Launch OpenCode in the folder where you want help with a real coding task.
Native OpenCode provider picker
Open the provider screen, select Xpersona, paste your setup key, pick xpersona/claude-fable-5, and run a real repo prompt.
No manual config required
The normal path is visual: open the provider screen, choose Xpersona, paste the setup key, pick the model, and start your prompt.
Launch OpenCode in the folder where you want help with a real coding task.
Open the provider screen, select Xpersona, and paste your Xpersona setup key.
Select xpersona/claude-fable-5. New Xpersona models appear in the same model picker as they are added.
Ask for a repo task. Successful requests show up in your Xpersona usage dashboard.
Model picker
Claude Fable 5
Premium alias for extra-high reasoning in OpenCode
xpersona/claude-fable-5Context
1,000,000
Output
128,000
Setup
Paste key once
Monthly packages
Every package includes Claude Fable 5. Allowances reset each month with no overage billing.
Builder
$20 / month
A focused monthly lane for individual builders.
Choose BuilderPro
$100 / month
Five times Builder capacity for daily coding-agent work.
Choose ProStudio
$200 / month
The highest public allowance for launches and recurring runs.
Choose StudioPackage limits: Builder 3,500,000 tokens, Pro 17,500,000, Studio 70,000,000.
Start with the $20 Builder package for 3.5M monthly tokens, or keep the 24-hour PAYG trial.
Open the provider screen in OpenCode, choose Xpersona, and paste your setup key.
Choose xpersona/claude-fable-5 in the model picker, then ask for the first repo task.
OpenCode FAQ
Xpersona is built around a short setup loop: choose the provider, pick the model, run a real coding task, then check usage without leaving the same account.
Yes. Xpersona appears as a native OpenCode provider. Choose Xpersona from the provider screen, paste your setup key, and select xpersona/claude-fable-5 from the model list.
No for the normal setup path. Start in the OpenCode UI, choose Xpersona, paste the setup key, and pick a model from the model list.
Use the dashboard usage page to inspect requests, status, tokens, and spend after your coding runs.