> ## Documentation Index > Fetch the complete documentation index at: https://docs.lighton.ai/llms.txt > Use this file to discover all available pages before exploring further. # Update a managed ML model (instance admin) > Update a managed model. Applied as a delete+recreate under the hood (LiteLLM's native update endpoint is unreliable). Reserved to Sys Admin / Admin roles. ## OpenAPI ````yaml /api-reference/openapi-v3-instance.yaml patch /api/v3/instance/ml-models/{uuid} openapi: 3.1.0 info: title: LightOn API version: 3.15.0 (v3-instance) description: >- LightOn gives you an API to search, parse, and ingest documents at scale. Build knowledge-retrieval pipelines without managing vector databases or OCR models. servers: - url: https://paradigm.lighton.ai security: [] tags: - name: Models description: Operations about AI models - name: Workspaces description: Operations about workspaces - name: Users description: Operations about users - name: User Groups description: Operations about user groups - name: Companies description: Operations about companies paths: /api/v3/instance/ml-models/{uuid}: patch: tags: - Models summary: Update a managed ML model (instance admin) description: >- Update a managed model. Applied as a delete+recreate under the hood (LiteLLM's native update endpoint is unreliable). Reserved to Sys Admin / Admin roles. operationId: api_v3_instance_ml_models_partial_update parameters: - in: path name: uuid schema: type: string format: uuid required: true requestBody: content: application/json: schema: $ref: '#/components/schemas/PatchedMLModelUpdatePayload' examples: ManagedModel: value: id: d612ede5-16b6-4d7b-b01b-5afb873b26bb name: alfred-ft5 technical_name: alfred-ft5 litellm_model: hosted_vllm/alfred-ft5 model_type: Large Language Model endpoint: http://models-gra11-afe6f177.lighton.ai:30424/v1/ max_model_len: 32768 deployment_type: Self-Hosted enabled: true is_default: false is_parser: false is_default_reranker: false use_vllm_guided_choice: true start_messages_template: null health: status: healthy is_healthy: true last_checked_at: '2026-08-12T09:31:44.512000Z' summary: Managed model application/x-www-form-urlencoded: schema: $ref: '#/components/schemas/PatchedMLModelUpdatePayload' multipart/form-data: schema: $ref: '#/components/schemas/PatchedMLModelUpdatePayload' responses: '200': content: application/json: schema: $ref: '#/components/schemas/MLModelAdminDetail' examples: ManagedModel: value: id: d612ede5-16b6-4d7b-b01b-5afb873b26bb name: alfred-ft5 technical_name: alfred-ft5 litellm_model: hosted_vllm/alfred-ft5 model_type: Large Language Model endpoint: http://models-gra11-afe6f177.lighton.ai:30424/v1/ max_model_len: 32768 deployment_type: Self-Hosted enabled: true is_default: false is_parser: false is_default_reranker: false use_vllm_guided_choice: true start_messages_template: null health: status: healthy is_healthy: true last_checked_at: '2026-08-12T09:31:44.512000Z' summary: Managed model description: '' '403': content: application/json: schema: $ref: '#/components/schemas/APIV3ErrorResponse' examples: Forbidden: value: detail: Instance admin access required. description: >- Instance admin access required. Only Sys Admin and Admin roles can manage the shared model catalog. '404': content: application/json: schema: $ref: '#/components/schemas/APIV3ErrorResponse' examples: NotFound: value: detail: ML model d612ede5-16b6-4d7b-b01b-5afb873b26bb not found. summary: Not found description: Managed model not found. '503': description: >- API is under maintenance. Check `GET /api/v3/system/status` for active periods and retry. content: application/json: schema: $ref: '#/components/schemas/ServiceMaintenance503' security: - bearerAuth: [] components: schemas: PatchedMLModelUpdatePayload: type: object properties: name: type: - string - 'null' description: Human-friendly display name. technical_name: type: - string - 'null' description: Unique technical identifier used for routing. litellm_model: type: - string - 'null' description: >- Underlying LiteLLM provider routing string (e.g. hosted_vllm/my-model). Distinct from technical_name — omit to keep the current value. model_type: type: - string - 'null' description: Model type (e.g. Large Language Model, Embedding Model). endpoint: type: - string - 'null' description: Model serving endpoint URL. max_model_len: type: - integer - 'null' description: Maximum context length in tokens. enabled: type: - boolean - 'null' description: Whether the model is active. default: type: - boolean - 'null' description: Whether this is the default generation model. parser: type: - boolean - 'null' description: Whether the model can be used as a parser. is_default_reranker: type: - boolean - 'null' description: Whether this is the default reranker. deployment_type: type: - string - 'null' description: Deployment type (Self-Hosted or External). use_vllm_guided_choice: type: - boolean - 'null' description: Use vLLM guided_choice for constrained decoding. MLModelAdminDetail: type: object properties: id: type: string format: uuid description: Unique ID for this model. name: type: string description: Display name. technical_name: type: string description: Instance-wide unique technical name used for routing. litellm_model: type: - string - 'null' description: >- Underlying LiteLLM provider routing string (e.g. hosted_vllm/my-model). model_type: type: string description: Model type. endpoint: type: - string - 'null' description: Model serving endpoint URL. max_model_len: type: - integer - 'null' description: Maximum context length in tokens. deployment_type: type: - string - 'null' description: Deployment type (Self-Hosted or External). enabled: type: boolean description: Whether the model is active instance-wide. is_default: type: boolean description: Whether this is the platform default generation model. is_parser: type: boolean description: Whether the model can be used as a parser. is_default_reranker: type: boolean description: Whether this is the default reranker. use_vllm_guided_choice: type: boolean description: Use vLLM guided_choice for constrained decoding. start_messages_template: type: - string - 'null' description: Start messages template name. health: allOf: - $ref: '#/components/schemas/MLModelHealth' description: >- Health of the model as of the last periodic gateway healthcheck. Read from cache — reading this endpoint never triggers a check. required: - enabled - health - id - is_default - is_default_reranker - is_parser - model_type - name - technical_name - use_vllm_guided_choice APIV3ErrorResponse: type: object properties: id: type: - string - 'null' description: >- Job/resource id when one already exists (useful for async error diagnosis); null otherwise. code: type: integer description: HTTP status code error: type: string description: Error code used by the UI as a translation key detail: type: string description: Human-readable error message for developers doc_url: type: string description: Link to the error-code documentation page required: - code - detail - doc_url - error - id ServiceMaintenance503: type: object description: >- Returned by the maintenance middleware when the requested endpoint is blocked. required: - detail - error - mode properties: detail: type: string example: System is under maintenance. error: type: string example: service_maintenance mode: type: string enum: - full_shutdown - warning_banner description: >- `full_shutdown` blocks all traffic; `warning_banner` also blocks and shows a dismissible toast. reason: type: string description: Operator-supplied maintenance reason, if any. started_at: type: string format: date-time endpoint_category_names: type: array items: type: string description: >- Non-empty only for category-scoped periods. Empty means all endpoints are affected. MLModelHealth: type: object properties: status: allOf: - $ref: '#/components/schemas/MLModelHealthStatusEnum' description: >- Health status of the model as of the last periodic healthcheck: 'healthy', 'unhealthy', or 'unknown' when the model has never been checked. Never triggers a check. * `healthy` - healthy * `unhealthy` - unhealthy * `unknown` - unknown is_healthy: type: boolean description: >- Whether the model is considered usable for routing. Optimistic: true for a model that has never been checked. Read status to tell 'healthy' apart from 'unknown'. last_checked_at: type: - string - 'null' format: date-time description: >- Timestamp of the last health check, or null when the model has never been checked required: - is_healthy - last_checked_at - status MLModelHealthStatusEnum: enum: - healthy - unhealthy - unknown type: string description: |- * `healthy` - healthy * `unhealthy` - unhealthy * `unknown` - unknown securitySchemes: bearerAuth: type: http scheme: bearer description: >- Bearer authentication header of the form `Bearer `, where `` is your auth token. ````