> ## Documentation Index > Fetch the complete documentation index at: https://docs.onvy.health/llms.txt > Use this file to discover all available pages before exploring further. # Create AI summary > Create a new AI-generated summary. The public create surface currently accepts `type=meal`; the resulting `MealSummary` is retrievable through the list and detail routes alongside system-generated summary types. For meal records that should also appear in the user's meals timeline and emit `meals:*` webhooks, use `POST /users/{user_id}/nutrition/meals` instead. That endpoint creates the meal, runs the same analysis, and links the resulting `MealSummary` via `summary_id`. ## OpenAPI ````yaml /api-reference/openapi.yaml post /users/{user_id}/ai/summaries openapi: 3.1.0 info: title: ONVY Health API description: >- AI-powered holistic health services API with data merging and metrics calculation version: 1.0.0 contact: name: ONVY Support email: support@onvy.health url: https://onvy.health license: name: Proprietary url: https://onvy.health/terms servers: - url: https://api.onvy.health description: Public API security: - bearerAuth: [] tags: - name: static description: Static configuration endpoints - name: users description: User management - name: baselines description: User baseline metrics - name: daily-records description: Daily health records - name: chat description: AI chat completions - name: facts description: User facts for individualization - name: activities description: User activities (workouts and mindfulness sessions) - name: ai-summaries description: | AI-generated summaries. This is how ONVY exposes domain-specific insights such as meal nutrition analysis, sleep insights, workouts, and daily, weekly, nutrition, trend, and impact analysis. Most summary types are system-generated; the public create surface currently accepts `type=meal`. - name: custom-records description: Custom records and types - name: meals description: Nutrition meal tracking paths: /users/{user_id}/ai/summaries: post: tags: - ai-summaries summary: Create AI summary description: | Create a new AI-generated summary. The public create surface currently accepts `type=meal`; the resulting `MealSummary` is retrievable through the list and detail routes alongside system-generated summary types. For meal records that should also appear in the user's meals timeline and emit `meals:*` webhooks, use `POST /users/{user_id}/nutrition/meals` instead. That endpoint creates the meal, runs the same analysis, and links the resulting `MealSummary` via `summary_id`. operationId: createAISummary parameters: - $ref: '#/components/parameters/UserIdParam' requestBody: description: AI summary creation data required: true content: application/json: schema: $ref: '#/components/schemas/CreateAISummaryRequest' examples: text_meal: summary: Text-based meal summary value: type: meal data: description: Grilled chicken breast with quinoa and steamed broccoli meal_time: '2024-10-28T12:00:00Z' portion_multiplier: 1.5 comment: Homemade healthy lunch image_meal: summary: Image-based meal summary value: type: meal data: image_base64: data:image/jpeg;base64,/9j/4AAQSkZJRg... meal_time: '2024-10-28T12:00:00Z' portion_multiplier: 1 comment: Delicious homemade pasta responses: '201': description: AI summary created successfully content: application/json: schema: $ref: '#/components/schemas/AISummary' '400': $ref: '#/components/responses/BadRequest' '500': $ref: '#/components/responses/InternalError' components: parameters: UserIdParam: name: user_id in: path required: true description: User ID (e.g., google_104940819145861640201) schema: type: string schemas: CreateAISummaryRequest: title: CreateAISummaryRequest type: object required: - type - data properties: type: type: string enum: - meal description: Type of AI summary to create. Currently only 'meal' is supported. data: description: Input data for the AI summary, structure depends on the type oneOf: - title: MealInput description: Input schema for creating or updating a meal type: object required: - meal_time properties: meal_time: type: string format: date-time description: ISO 8601 timestamp of when the meal was consumed example: '2024-10-28T12:00:00Z' description: type: string description: >- Text description of the meal/food consumed. Required if image_base64 is not provided. maxLength: 2000 example: Grilled chicken breast with quinoa and steamed broccoli image_base64: type: string description: >- Base64-encoded image data with data URI prefix. Required if description is not provided. pattern: ^data:image/(jpeg|png|jpg);base64,[A-Za-z0-9+/=]+$ example: data:image/jpeg;base64,/9j/4AAQSkZJRg... portion_multiplier: type: number description: Multiplier for portion size (default 1.0) default: 1 minimum: 0.1 maximum: 10 example: 1.5 comment: type: string description: Additional comment about the meal maxLength: 500 example: Homemade healthy lunch anyOf: - required: - description - required: - image_base64 AISummary: title: AISummary type: object required: - id - name - user_id - project_id - created_at - updated_at - period - data - llm_output properties: id: type: string description: Unique summary identifier name: type: string enum: - WorkoutSummary - MindfulsessionSummary - MealSummary - SleepSummary - DailySummary - WeeklySummary - TrendSummary - ImpactSummary - WeeklyActivitySummary - DailyNutritionSummary - WeeklyNutritionSummary user_id: type: string project_id: type: string created_at: type: string format: date-time updated_at: type: string format: date-time providers: type: array items: type: string enum: - GARMIN - APPLE - OURA - FITBIT - WHOOP is_custom: type: boolean description: If provided by customer POST ref: type: string description: Optional reference to original data period: type: object required: - type - id - start - end properties: type: type: string enum: - activity - day - week - month - quarter - half-year - year id: type: string description: Period identifier (e.g., 2024-W43, 2024-10-21) start: type: string format: date-time end: type: string format: date-time duration_seconds: type: integer description: Duration in seconds (for activity periods) data: type: object description: Additional metrics specific to summary type llm_output: type: object required: - language - title - text properties: language: type: string description: Language code (e.g., en, de) title: type: string description: Generated title text: type: string description: Generated summary text followup_questions: type: array items: type: string knowledge_urls: type: array items: type: object required: - title - url properties: title: type: string url: type: string format: uri user_comment: type: - string - 'null' user_rating: type: integer enum: - -1 - 0 - 1 description: '-1 = bad, 0 = no rating, 1 = good' llm_debug: type: object properties: model: type: string description: Model used for generation prompt_id: type: string description: Prompt template identifier org_prompt_id: type: string description: Organization-specific prompt override ValidationError: type: object title: ValidationError required: - error properties: error: type: string minLength: 1 additionalProperties: false responses: BadRequest: description: Bad request content: application/json: schema: $ref: '#/components/schemas/ValidationError' InternalError: description: Internal server error content: application/json: schema: $ref: '#/components/schemas/ValidationError' securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT description: OAuth bearer token for Health API authentication ````