windmill-labs/windmill · error

No code completion model selected

Error message

No code completion model selected

What it means

Thrown by autocompleteRequest before any HTTP call when `copilotInfo.codeCompletionModel` is unset — the instance or workspace has no code-completion (FIM) model configured, so there is nothing to send the prefix/suffix context to. Configuration is missing at the deployment level, not in the calling editor.

Source

Thrown at frontend/src/lib/components/copilot/autocomplete/request.ts:30

		.map((line) => `${commentSymbol} ${line}`)
		.join('\n')
}

export async function autocompleteRequest(
	context: {
		prefix: string
		suffix: string
		scriptLang: ScriptLang | 'bunnative' | 'jsx' | 'tsx' | 'json'
		markers: editor.IMarker[]
		libraries: string
		workflowAsCode?: boolean
	},
	abortController: AbortController
) {
	const providerModel = get(copilotInfo).codeCompletionModel

	if (!providerModel) {
		throw new Error('No code completion model selected')
	}

	// Only add context lines for native FIM providers - other providers use chat completion
	// too much context degrades significantly the quality of the completion
	if (providerModel.provider === 'mistral' || providerModel.provider === 'deepseek') {
		let commentSymbol = getCommentSymbol(context.scriptLang)
		let contextLines = comment(
			commentSymbol,
			'You are a code completion assistant. You are given three important contexts (<LANGUAGE CONTEXT>, <DIAGNOSTICS>, <LIBRARY METHODS>) to help you complete the code.\n'
		)
		contextLines += comment(commentSymbol, 'LANGUAGE CONTEXT:\n')
		contextLines += comment(
			commentSymbol,
			getLangContext(context.scriptLang, { workflowAsCode: context.workflowAsCode }) + '\n'
		)
		contextLines += comment(commentSymbol, 'DIAGNOSTICS:\n')
		contextLines += comment(commentSymbol, context.markers.map((m) => m.message).join('\n') + '\n')
		contextLines += comment(commentSymbol, 'LIBRARY METHODS:\n')

View on GitHub (pinned to e474e8803c)

Solutions

1. Select a code completion model in the instance or workspace AI settings
2. Verify the copilot info endpoint returns a non-empty codeCompletionModel
3. Confirm the editor only invokes autocomplete after copilot info has loaded

When it happens

Trigger: Thrown at frontend/src/lib/components/copilot/autocomplete/request.ts:30 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of windmill-labs/windmill@e474e8803c (2026-09-03). Data as JSON: /api/errors/03bc75b606964557. Report an issue: GitHub.