What causes "model.yml not found for ${modelId}" in janhq/jan?

Confirm the modelId is correct and belongs to the llamacpp provider.

model.yml not found for

Error message

model.yml not found for ${modelId}

What it means

Thrown by the model-configuration read path when models/<modelId>/model.yml cannot be found. This YAML file holds the model's config (context length, mtp/speculative-decoding params, etc.), so without it the extension cannot build or validate the model's runtime settings.

Solutions

Confirm the modelId is correct and belongs to the llamacpp provider.
Re-import or re-download the model so model.yml is regenerated.
Check the provider's models/ directory on disk to see what actually exists.
Refresh the model list / restart the app to clear stale catalog entries.

Example fix

// before
const cfg = await getModelConfig(modelId)
// after
const models = await extension.listModels()
if (!models.some(m => m.id === modelId)) {
  throw new Error(`Model ${modelId} is not registered with the llamacpp provider`)
}
const cfg = await getModelConfig(modelId)

Defensive patterns

Strategy: validation

Validate before calling

const settings = await listModelSettings()
if (!settings.some(s => s.modelId === modelId)) {
  throw new Error(`No llamacpp settings exist for ${modelId}`)
}

Type guard

null

Try / catch

try {
  const cfg = await getModelConfig(modelId)
} catch (e) {
  if (String(e).includes('model.yml not found')) {
    promptUserToReimport(modelId)
    return defaultModelConfig
  }
  throw e
}

Prevention

Import models only through the extension so model.yml is generated.
Verify the modelId belongs to the llamacpp provider, not another provider.
Re-download interrupted imports before reading settings.
Refresh the catalog after any out-of-band file deletion.

When it happens

Trigger: Reading model settings/options for a modelId whose model.yml is missing — model not imported through this extension, import interrupted, files manually deleted, or modelId typo/wrong casing.

Common situations: Opening model settings in Jan UI for a model from another provider; a failed download left the directory without model.yml; the model was deleted while its settings page was open.

Understand the failure class

Background: "Config file not found": what it means and how to fix it in docker-sync, Maven, Vagrant, Turborepo and other tools — this error's family across 60 libraries.

Related errors

AI-assisted analysis of janhq/jan@7205d770c1 (2026-09-17). Data as JSON: /api/errors/42a74608cd081351. Report an issue: GitHub.

Appendix: source

Thrown at extensions/llamacpp-extension/src/index.ts:2661

  }

  async updateMtpSettings(
    modelId: string,
    patch: {
      mtp?: boolean
      spec_draft_n_max?: number | null
      spec_draft_n_min?: number | null
      spec_draft_p_min?: number | null
    }
  ): Promise<void> {
    const configPath = await joinPath([
      await this.getProviderPath(),
      'models',
      modelId,
      'model.yml',
    ])
    if (!(await fs.existsSync(configPath))) {
      throw new Error(`model.yml not found for ${modelId}`)
    }
    const cfg = (await invoke<ModelConfig>('read_yaml', { path: configPath })) as ModelConfig & {
      mtp?: boolean
      spec_draft_n_max?: number
      spec_draft_n_min?: number
      spec_draft_p_min?: number
    }

    if (typeof patch.mtp === 'boolean') cfg.mtp = patch.mtp
    const assignNumeric = (
      key: 'spec_draft_n_max' | 'spec_draft_n_min' | 'spec_draft_p_min',
      value: number | null | undefined
    ) => {
      if (value === null) {
        delete cfg[key]
      } else if (typeof value === 'number' && Number.isFinite(value)) {
        cfg[key] = value
      }

View on GitHub (pinned to 7205d770c1)