Use PaletteAI Inference Launchpad with OpenCode
This guide explains how to connect OpenCode to a PaletteAI Inference Launchpad appliance so that a model running on the appliance serves every request instead of a cloud provider. You add a custom provider to the OpenCode configuration file and confirm the connection.
Prerequisites
- OpenCode installed and already working. For installation details, refer to the OpenCode website.
- A running PaletteAI Inference Launchpad appliance with at least one model deployed and serving. To deploy a model, refer to Deploy a Model.
- An API token for the appliance. To create one, refer to Generate an API Token, or use a token an administrator generated for you.
Configure OpenCode
OpenCode connects to any OpenAI-compatible endpoint through a custom provider. Add a provider for the appliance to the OpenCode configuration file. For a description of each field, refer to OpenCode Configuration.
The console can generate a starter version of this file for you. Select Connect coding agent and open the
OpenCode tab to copy an opencode.json snippet pre-filled with your appliance's endpoint. Review the baseURL and
models values against the steps below before you save it.
-
Add the following provider to the OpenCode configuration file at
~/.config/opencode/opencode.json. Replace<appliance-host>with your appliance address and<lpai-token>with the token you copied.{
"$schema": "https://opencode.ai/config.json",
"provider": {
"launchpad": {
"npm": "@ai-sdk/openai-compatible",
"name": "Launchpad",
"options": {
"baseURL": "https://<appliance-host>/v1",
"apiKey": "<lpai-token>"
},
"models": {
"glm-5.2": { "name": "GLM-5.2 (Launchpad)" }
}
}
}
} -
Set
baseURLto your appliance address with the/v1path appended. -
Set
apiKeyto yourlpai_token. -
Under
models, list each model id the appliance serves that you want to use, such asglm-5.2. Thelaunchpadkey is a name you choose for the provider. OpenCode identifies a model by that provider name and a model id joined with a slash, such aslaunchpad/glm-5.2, which you pass to the--modelflag when you run OpenCode in the next section. -
We strongly recommend giving the appliance a DNS name and a valid, publicly trusted TLS certificate. The connection uses HTTPS, so a valid certificate protects your token in transit.
warningIf the appliance uses a self-signed certificate, OpenCode rejects the connection by default because it runs on Node.js. As a temporary measure for testing, set
NODE_TLS_REJECT_UNAUTHORIZED=0before you start OpenCode. This disables certificate verification, so do not use it outside short-lived testing.
Verify the Connection
Run a single prompt to confirm the appliance answers. The --model flag takes a provider/model value that combines
the provider key from your configuration file with a model id.
opencode run --model launchpad/glm-5.2 "reply with exactly OPENCODE_OK"
OPENCODE_OK
A reply confirms that the base URL, token, provider, and model routing all work. OpenCode splits the --model value on
the first slash, so launchpad/glm-5.2 selects the glm-5.2 model from the launchpad provider.
If you use a reasoning model and the reply comes back empty, raise the output limit. Hidden reasoning tokens can consume a small output budget entirely, which leaves no room for the visible reply.
Request Routing and Quotas
Requests you send through the appliance are subject to the routing rule and quota configured for your client. If an
operator configured a routing rule for your token, the appliance can redirect a request to a frontier model instead of a
local one. Quotas apply to the client, so every API token that belongs to the client draws on the same limits. When the
client exhausts a quota, the appliance returns an HTTP 429 response.
To understand what a client is and how API keys and quotas govern usage, refer to Clients and Quotas. To understand how the Tier map and the semantic router combine to pick a model for each request, refer to Routing Behavior.
Next Steps
To look up each configuration value, refer to OpenCode Configuration. To connect a different coding tool, refer to Use PaletteAI Inference Launchpad with Claude Code, Use PaletteAI Inference Launchpad with Cursor, or Use PaletteAI Inference Launchpad with OpenAI Codex. To deploy another model to the appliance, refer to Deploy a Model.