⌁ Guides

Zot

Gondola ships as a built-in provider in Zot, the single-binary Go coding agent. One env var and you are running any Venice model, paid per request in USDC on Base.

Install Zot

Gondola is built in from v0.3.43 onward, so check your version if you already have the binary.

macOS, Linux

curl -fsSL https://www.zot.sh/install.sh | bash
zot --version

Windows, PowerShell

iwr -useb https://www.zot.sh/install.ps1 | iex

Create an API key

Sign in with your Base wallet at /dashboard?tab=keys and mint a Gondola API key. Keys start with gnd_. It authenticates every request and each call is charged in USDC against your Gondola balance.

Top up USDC on Base once at /wallet. The same balance powers the browser chat and every tool you connect.

Run it

Zot derives the env var from the provider id, upper snake case, so gondola reads GONDOLA_API_KEY. No config file: the provider already points at https://api.gondola-ai.com/v1.

shell

export GONDOLA_API_KEY=gnd_xxxx
zot --provider gondola

Other ways to supply the key

The env var isn't the only path: run zot and type /login to paste the key interactively, or pass --api-key for a one-off run. Zot checks the key against Gondola's authenticated balance endpoint at login rather than waiting for the first failed call, so a bad key or an empty balance surfaces immediately.

Pick a model

The default is kimi-k3, a strong agentic coder at a fraction of frontier pricing. Switch any time with /model inside the TUI, or --model <id> on the command line.

Zot discovers Gondola's live text catalog in the background and adds it to the picker, so a model Venice adds today is selectable the same day with no config edit. Current prices are on the model board.

shell

zot --list-models
zot --provider gondola --model claude-sonnet-5

Prompt privacy

Gondola never stores prompts or completions: requests are metered, forwarded, and gone when the response finishes (the one exception, a video source clip over 18 MiB, is held privately for the length of its job; see /privacy), and Venice runs a no-logging model underneath. For zero retention on Venice-controlled infrastructure too, pick a model whose privacy field reads private on GET /v1/models. Details on /privacy.

Notes and troubleshooting

Cap spend at the key, not the request

Zot has no field for custom request headers, so x-gondola-max-factor and x-gondola-max-charge-usd aren't reachable from it. For unattended runs, cap spend at the source instead: mint a dedicated key with a spend limit at /dashboard?tab=keys, editable any time.

If a request returns insufficient balance, top up at /wallet. Every response also carries your remaining balance in x-gondola-balance.

Loading page