linshenkx/prompt-optimizer · error · Error

This model does not support text-to-image or image-to-image

Error message

This model does not support text-to-image or image-to-image generation

What it means

selectTestType in useImageModelManager derives which generation mode to test by checking the selected model's text2image/image2image capability flags. If neither capability is true, it throws because there is no valid image test to run for that model.

Source

Thrown at packages/ui/src/composables/model/useImageModelManager.ts:397

  // 辅助函数：根据模型能力选择测试类型
  const selectTestType = (model: ImageModel): 'text2image' | 'image2image' => {
    const capabilities = model.capabilities || {}
    const text2image = capabilities?.text2image
    const image2image = capabilities?.image2image

    if (text2image && !image2image) {
      return 'text2image'  // 只支持文生图
    }

    if (!text2image && image2image) {
      return 'image2image' // 只支持图生图
    }

    if (text2image && image2image) {
      return 'text2image'  // 两种都支持，优先文生图
    }

    throw new Error('This model does not support text-to-image or image-to-image generation')
  }

  const testConnection = async () => {
    if (!selectedProvider.value || !hasValidConnectionConfig(configForm.value.connectionConfig || {})) {
      return
    }

    // 检查是否选择了模型
    if (!configForm.value.modelId) {
      toast.error(t('image.model.selectRequired'))
      return
    }

    isTestingConnection.value = true
    connectionStatus.value = { type: 'info', messageKey: 'image.connection.testing' }

    try {
      // 本地先做详细校验，给出缺失与类型错误提示

View on GitHub (pinned to 3e677b1d9f)

Solutions

1. Filter or flag models in the picker so only models with text2image or image2image capability can reach the test flow.
2. Ensure capability detection (probe/adapter) runs and populates text2image/image2image before selectTestType is called.
3. Catch the error and show a capability warning instead of crashing the dialog.

Example fix

// before
const testType = selectTestType(model)

// after
const caps = model.capabilities ?? {}
if (!caps.text2image && !caps.image2image) {
  toast.warning('Selected model has no image generation capability')
  return null
}
const testType = selectTestType(model)

const caps = model?.capabilities ?? {}
const canTestImage = Boolean(caps.text2image || caps.image2image)
if (!canTestImage) { /* skip test, warn user */ }

const supportsImageGeneration = (m: { capabilities?: { text2image?: boolean; image2image?: boolean } }): boolean =>
  Boolean(m.capabilities?.text2image || m.capabilities?.image2image)

try { const t = selectTestType(model) } catch (e) { if (e instanceof Error && /does not support/.test(e.message)) { warnNoImageCapability(); return } throw e }

When it happens

Trigger: Calling testType/selectTestType for a model whose capabilities object has text2image=false (or missing) and image2image=false (or missing), e.g. an LLM-only or video-only model selected in the image model manager.

Common situations: User pastes a custom model ID that defaults to no image capabilities; provider returns a model listing where capability flags were not mapped; model metadata cached from an older version lacking the capability fields.

Related errors

• TEXT2IMAGE_INPUT_IMAGE_NOT_ALLOWED
• CONFIG_NOT_FOUND
• CONTEXT_ERROR_CODES.STORAGE_ERROR
• CONTEXT_ERROR_CODES.IMPORT_FORMAT_ERROR
• DATA_ERROR_CODES.ELECTRON_API_UNAVAILABLE

AI-assisted analysis of linshenkx/prompt-optimizer@3e677b1d9f (2026-08-27). Data as JSON: /api/errors/2edf9f07edfbbde0. Report an issue: GitHub.