> ## 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. # Create a custom model > Register a new custom ML model for the authenticated user's company. The litellm_model field is immutable after creation. For OpenAI-compatible endpoints (LM Studio, Ollama), prefix with 'openai/' (e.g. 'openai/llama3'). An optional temperature is applied to every request sent to this model, in place of whatever the calling feature would otherwise use. Some models accept exactly one temperature and are set to it regardless of what is submitted; the model catalogue reports that value as required_temperature. Requires company admin role. ## OpenAPI ````yaml /api-reference/openapi-v3-company.yaml post /api/v3/company/models openapi: 3.1.0 info: title: LightOn API version: 3.15.0 (v3-company) 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/company/models: post: tags: - Models summary: Create a custom model description: >- Register a new custom ML model for the authenticated user's company. The litellm_model field is immutable after creation. For OpenAI-compatible endpoints (LM Studio, Ollama), prefix with 'openai/' (e.g. 'openai/llama3'). An optional temperature is applied to every request sent to this model, in place of whatever the calling feature would otherwise use. Some models accept exactly one temperature and are set to it regardless of what is submitted; the model catalogue reports that value as required_temperature. Requires company admin role. operationId: api_v3_company_models_create requestBody: content: application/json: schema: $ref: '#/components/schemas/CustomMLModelCreateRequest' examples: LMStudioModel: value: name: Gemma 4 (local LM Studio) litellm_model: openai/google/gemma-4-e4b model_type: Large Language Model endpoint: http://host.docker.internal:1234/v1 api_key: null temperature: 0.2 summary: LM Studio model CustomModel: value: id: a08cff52-1a96-49f4-9f28-84a060b9264c name: Gemma 4 (local LM Studio) technical_name: custom-42-a08cff52-1a96-49f4-9f28-84a060b9264c litellm_model: openai/google/gemma-4-e4b model_type: Large Language Model endpoint: http://host.docker.internal:1234/v1 enabled: true is_default: false temperature: 0.2 required_temperature: null max_temperature: null summary: Custom model application/x-www-form-urlencoded: schema: $ref: '#/components/schemas/CustomMLModelCreateRequest' multipart/form-data: schema: $ref: '#/components/schemas/CustomMLModelCreateRequest' required: true responses: '201': content: application/json: schema: $ref: '#/components/schemas/CustomMLModelResponse' examples: LMStudioModel: value: name: Gemma 4 (local LM Studio) litellm_model: openai/google/gemma-4-e4b model_type: Large Language Model endpoint: http://host.docker.internal:1234/v1 api_key: null temperature: 0.2 summary: LM Studio model CustomModel: value: id: a08cff52-1a96-49f4-9f28-84a060b9264c name: Gemma 4 (local LM Studio) technical_name: custom-42-a08cff52-1a96-49f4-9f28-84a060b9264c litellm_model: openai/google/gemma-4-e4b model_type: Large Language Model endpoint: http://host.docker.internal:1234/v1 enabled: true is_default: false temperature: 0.2 required_temperature: null max_temperature: null summary: Custom model description: '' '400': content: application/json: schema: $ref: '#/components/schemas/APIV3ErrorResponse' examples: AboveTheProviderMaximum: value: id: null code: 400 error: bad_request detail: >- Temperature 1.5 is above the maximum of 1.0 accepted by this model's provider. doc_url: https://developers.lighton.ai/errors#bad_request summary: Above the provider maximum CustomModelsUnavailableOnThisDeployment: value: id: null code: 400 error: bad_request detail: >- This deployment cannot serve company custom models: they are only available when the LiteLLM gateway is the configured model provider. doc_url: https://developers.lighton.ai/errors#bad_request summary: Custom models unavailable on this deployment description: >- Either the requested temperature is above the maximum the model's provider accepts (see the update endpoint for the per-provider ceilings), or the deployment cannot serve company custom models at all. The second is only possible here, on registration: it depends on the configured model provider, and the same request would succeed on a deployment using the LiteLLM gateway. Listing and deleting stay available either way, so models registered before a provider change can still be cleaned up. '403': content: application/json: schema: $ref: '#/components/schemas/APIV3ErrorResponse' examples: Forbidden: value: id: null code: 403 error: insufficient_permissions detail: Admin access required. doc_url: >- https://developers.lighton.ai/errors#insufficient_permissions description: >- Admin access required. Only company admins and superadmins can write custom models. '422': content: application/json: schema: $ref: '#/components/schemas/APIV3ErrorResponse' examples: MissingRequiredField: value: id: null code: 422 error: validation_error detail: One or more fields failed validation. doc_url: https://developers.lighton.ai/errors#validation_error fields: litellm_model: - error: required detail: Field required summary: Missing required field DuplicateName: value: id: null code: 422 error: validation_error detail: One or more fields failed validation. doc_url: https://developers.lighton.ai/errors#validation_error fields: name: - error: already_exists detail: >- A custom model named 'My GPT-4' already exists for this company. summary: Duplicate name description: >- A required field is missing, a field holds a value of the wrong type, or the name collides with an existing custom model in the same company. '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: CustomMLModelCreateRequest: type: object properties: name: type: string description: Display name, unique within the company. litellm_model: type: string description: >- LiteLLM model string, e.g. 'openai/gpt-4-turbo' or 'azure/my-deployment'. model_type: type: string default: Large Language Model description: Model type (e.g. 'Large Language Model', 'Embedding Model'). endpoint: type: - string - 'null' description: Custom API base URL (api_base in LiteLLM). api_key: type: - string - 'null' description: API key for the custom model endpoint. temperature: type: - number - 'null' format: double maximum: 2 minimum: 0 description: >- Sampling temperature to use for every request to this model. Omit to let each calling feature use its own value. Models that accept only one temperature are set to it regardless of what is submitted here. required: - litellm_model - name CustomMLModelResponse: type: object properties: id: type: string format: uuid description: Unique ID for this custom model. name: type: string description: Display name, unique within the company. technical_name: type: string description: LiteLLM identifier (custom-{uuid}). litellm_model: type: string description: LiteLLM model string, e.g. 'openai/gpt-4-turbo'. model_type: type: string description: Model type. endpoint: type: - string - 'null' description: Custom API base URL. enabled: type: boolean description: Whether the model is active. is_default: type: boolean description: Whether this is the company's default custom model. temperature: type: - number - 'null' format: double description: >- Sampling temperature sent for every request to this model. Null means no temperature is configured and the calling feature's own value is used. required_temperature: type: - number - 'null' format: double description: >- The only temperature this model accepts, when it accepts exactly one. Set for models that have removed sampling parameters; null for every other model. max_temperature: type: - number - 'null' format: double description: >- Highest temperature this model's provider accepts. Null for a model served from its own endpoint, where no provider bound applies. required: - enabled - id - is_default - litellm_model - model_type - name - technical_name 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. securitySchemes: bearerAuth: type: http scheme: bearer description: >- Bearer authentication header of the form `Bearer `, where `` is your auth token. ````