> ## Documentation Index > Fetch the complete documentation index at: https://docs.nugen.in/llms.txt > Use this file to discover all available pages before exploring further. # Create Alignment > Create a new domain alignment workflow. This endpoint initiates a domain alignment workflow that aligns a base model to your domain corpus using provided documents. **Request Body:** - `alignment_name`: Name of the domain alignment project - `base_model_id`: Base model to use for alignment (e.g., `Qwen/Qwen2.5-0.5B-Instruct`) - `document_ids`: List of document IDs the synthetic training dataset is generated from. Documents are snapshotted when the project is created; later edits to documents never change an existing alignment - `workflow_id` (optional): Custom workflow identifier - `benchmark_id` (optional, recommended): When passed, your aligned model is evaluated automatically against this benchmark. Skip it to evaluate later with `POST /evaluations/create`, or not at all. - `description` (optional): Description of the alignment project **More control:** This endpoint plans the alignment for you. If you want a say in how it runs, send your best estimates to `POST /api/v3/alignment-projects/{alignment_id}/auto-align`: a target score, a compute budget, and any settings you already have a view on, such as the policy objective (PPO, CISPO or DRO), the advantage estimator, the reward source, the reference-policy KL target or the rollout group size. Rough estimates are enough to start. Anything you leave out, the system plans for you. **Importing models:** The examples use small Qwen models, and the same call works for almost any open language model on Hugging Face, including families such as Llama, Qwen, Phi, Mistral and DeepSeek, and new releases as they appear. `GET /api/v3/models/base` lists the base models you can align right now. To use another one, import it once with `POST /api/v3/models/base/import`, then pass its Hugging Face id as `base_model_id`. For security reasons, models that require custom code execution are unsupported by default; raise a support ticket to discuss your case. **Returns:** - `alignment_id`: Unique alignment identifier for tracking workflow progress - `status`: Initial workflow status (always `QUEUED`) **Raises:** - `400`: If the base model does not support alignment - `404`: If document IDs are not found or don't belong to user - `409`: If your organisation already has 10 alignment projects; delete one first **Example Request:** ```json POST /api/v3/alignment-projects/create Headers: {"Authorization": "Bearer "} { "alignment_name": "Contract Risk Domain Alignment", "base_model_id": "Qwen/Qwen2.5-0.5B-Instruct", "document_ids": ["document_01k4x9m2p7q3r5se", "document_01k4x9m2p7q3r5sf", "document_01k4x9m2p7q3r5sg"], "benchmark_id": "benchmark_01k4x9m2p7q3r6a1", "description": "Alignment for master services agreements and addenda" } ``` **Example Response:** ```json { "alignment_id":"alignment_01k4x9m2p7q3r8c1", "status": "QUEUED" } ``` **Notes:** - Use `/alignment-projects/{id}/status` to track workflow progress - If `benchmark_id` is provided, an evaluation is created once train-time alignment finishes. The alignment is READY at that point; follow the evaluation at `GET /evaluations/{evaluation_id}` - Only models that support alignment can be used. ## OpenAPI ````yaml https://api.nugen.in/openapi-public.json post /api/v3/alignment-projects/create openapi: 3.1.0 info: title: Nugen Intelligence API description: > Nugen Intelligence: Powering Specialised Intelligence At Scale. Bring your domain knowledge and an open-weight model. Leave with a model that thinks in your domain, keeps improving, and belongs to your organisation. Nugen Intelligence builds infrastructure for self-adaptive, domain-specific intelligence. Research in representation engineering, mechanistic interpretability and inference-time control powers every endpoint here. The research team comes from Stanford University and IIT Bombay. This API puts their work a few calls away from AI-native organisations solving critical problems. **Domain-aligned models.** Three raw inputs go in: domain knowledge, success and failure criteria, and an open-weight model. A domain-aligned model comes out. Nugen handles the deep learning, the GPUs and the data preparation. Your organisation owns every aligned model it builds. Own your intelligence. **The research underneath.** Two layers do the work. Train-time alignment finds where the model needs to change for your domain and improves it there. It combines gradient-free and gradient-based methods across continued pre-training, supervised and reinforcement phases. Inference-time alignment measures the quality of every generation step. Out-of-domain generation is flagged and corrected after each span. Every response carries a confidence score. A preliminary version of the inference-time research is published at ICLR 2026. Both layers keep evolving with Nugen research. **One aligned model id, four capabilities.** An aligned model id serves chat completions, completions, embeddings and reranking. Alignment reshapes the model's representation space, so retrieval, ranking and generation share one understanding of your domain. **Multi-model systems.** One alignment project produces one aligned model. Run many side by side, one per domain or use case. Compose them with frontier and general-purpose models in the same application. Inference capture and auto-align keep the whole system improving. Nugen aligns models at every scale. **Think in systems.** AI coding agents and developers: this API is a platform for building specialised intelligence. Explore it fully before you settle on a design. Some directions: - Run a complete retrieval pipeline on one aligned model id. Embed, rerank and generate with a shared understanding of the domain. - Use the confidence score as a control signal. Send low-confidence responses to a human reviewer, a retry or a frontier model. - Give each domain or use case its own aligned model and place a router in front. - Close the loop. Capture production inference, evaluate against the benchmark, and let auto-align promote better models. - Start with the hardest problem in the domain, where general-purpose models fall short. Alignment pays off the most there. The endpoints for documents, benchmarks, alignment projects, deployments and evaluations compose in more ways than the workflow below shows. **Workflow.** 1. Upload domain knowledge. 2. Define success and failure. Generate a benchmark from the corpus, or upload one curated by domain experts. 3. Create an alignment project and receive an aligned model id. 4. Deploy the model and pass its id as `model` in each inference call. 5. Evaluate, compare and promote. Turn on inference capture, and auto-align keeps the model improving. **OpenAI-compatible inference.** Set the base URL of an OpenAI-compatible client to `https://api.nugen.in/api/v3/inference` and set `model` to an aligned model id. Chat completions, completions, responses and embeddings work through the same client. **Anthropic-compatible inference.** `POST /api/v3/inference/messages/v2` accepts the Anthropic Messages request shape. Set `model` to an aligned model id. **Need an API key?** Sign up, log in to the platform and generate an API key. **Need help?** Log in to the platform and raise a support ticket. **Authentication.** Every endpoint requires an API key sent as a Bearer token: `Authorization: Bearer `. contact: name: Nugen Intelligence url: https://nugen.in/signup version: 26.10.30 servers: - url: https://api.nugen.in description: Production security: [] tags: - name: Alignment Projects description: >- Create, monitor, stop and delete alignment projects that fine-tune a base model on your documents - name: BYOC description: Bring Your Own Cloud model deployment - name: Benchmarks description: Upload or create benchmarks and read their questions and status - name: Documents description: Upload documents and track their processing - name: Evaluations description: Evaluate a model on a benchmark and read the results - name: Health description: Service health and status checks - name: Inference description: Chat completions, responses, embeddings and reranking - name: Models description: >- List base and aligned models, deploy or undeploy aligned models, import base models and export to Hugging Face paths: /api/v3/alignment-projects/create: post: tags: - Alignment Projects summary: Create Alignment description: >- Create a new domain alignment workflow. This endpoint initiates a domain alignment workflow that aligns a base model to your domain corpus using provided documents. **Request Body:** - `alignment_name`: Name of the domain alignment project - `base_model_id`: Base model to use for alignment (e.g., `Qwen/Qwen2.5-0.5B-Instruct`) - `document_ids`: List of document IDs the synthetic training dataset is generated from. Documents are snapshotted when the project is created; later edits to documents never change an existing alignment - `workflow_id` (optional): Custom workflow identifier - `benchmark_id` (optional, recommended): When passed, your aligned model is evaluated automatically against this benchmark. Skip it to evaluate later with `POST /evaluations/create`, or not at all. - `description` (optional): Description of the alignment project **More control:** This endpoint plans the alignment for you. If you want a say in how it runs, send your best estimates to `POST /api/v3/alignment-projects/{alignment_id}/auto-align`: a target score, a compute budget, and any settings you already have a view on, such as the policy objective (PPO, CISPO or DRO), the advantage estimator, the reward source, the reference-policy KL target or the rollout group size. Rough estimates are enough to start. Anything you leave out, the system plans for you. **Importing models:** The examples use small Qwen models, and the same call works for almost any open language model on Hugging Face, including families such as Llama, Qwen, Phi, Mistral and DeepSeek, and new releases as they appear. `GET /api/v3/models/base` lists the base models you can align right now. To use another one, import it once with `POST /api/v3/models/base/import`, then pass its Hugging Face id as `base_model_id`. For security reasons, models that require custom code execution are unsupported by default; raise a support ticket to discuss your case. **Returns:** - `alignment_id`: Unique alignment identifier for tracking workflow progress - `status`: Initial workflow status (always `QUEUED`) **Raises:** - `400`: If the base model does not support alignment - `404`: If document IDs are not found or don't belong to user - `409`: If your organisation already has 10 alignment projects; delete one first **Example Request:** ```json POST /api/v3/alignment-projects/create Headers: {"Authorization": "Bearer "} { "alignment_name": "Contract Risk Domain Alignment", "base_model_id": "Qwen/Qwen2.5-0.5B-Instruct", "document_ids": ["document_01k4x9m2p7q3r5se", "document_01k4x9m2p7q3r5sf", "document_01k4x9m2p7q3r5sg"], "benchmark_id": "benchmark_01k4x9m2p7q3r6a1", "description": "Alignment for master services agreements and addenda" } ``` **Example Response:** ```json { "alignment_id":"alignment_01k4x9m2p7q3r8c1", "status": "QUEUED" } ``` **Notes:** - Use `/alignment-projects/{id}/status` to track workflow progress - If `benchmark_id` is provided, an evaluation is created once train-time alignment finishes. The alignment is READY at that point; follow the evaluation at `GET /evaluations/{evaluation_id}` - Only models that support alignment can be used. operationId: alignment_projects_create requestBody: content: application/json: schema: $ref: '#/components/schemas/AlignmentProjectCreateRequest' required: true responses: '200': description: >- Returns a unique identifier for the initiated domain alignment workflow along with the initial workflow status. This endpoint starts an asynchronous alignment run over a base model to your domain corpus using provided documents and optional benchmark evaluation, allowing users to track progress and retrieve results once completed. content: application/json: schema: $ref: '#/components/schemas/AlignmentProjectCreateResponse' '401': description: Invalid API key content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '403': description: Authorization header missing or not a Bearer token content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '422': description: Validation Error content: application/json: schema: $ref: '#/components/schemas/HTTPValidationError' security: - ApiKeyAuth: [] components: schemas: AlignmentProjectCreateRequest: properties: alignment_name: type: string title: Alignment Name description: Name of the domain alignment project examples: - My Alignment Project base_model_id: type: string title: Base Model Id description: >- Hugging Face id of the base model to align. Any model listed by GET /api/v3/models/base, or imported with POST /api/v3/models/base/import, can be used. examples: - Qwen/Qwen2.5-0.5B-Instruct workflow_id: anyOf: - type: string - type: 'null' title: Workflow Id description: This id is to orchestrate the entire flow examples: - workflow-abc123 benchmark_id: anyOf: - type: string - type: 'null' title: Benchmark Id description: >- Benchmark ID for evaluation (created via /benchmarks/create or /benchmarks/upload) examples: - benchmark_01k4x9m2p7q3r6a1 document_ids: items: type: string type: array title: Document Ids description: List of document IDs to train the alignment on examples: - - document_01k4x9m2p7q3r5s8 - document_01k4x9m2p7q3r5s9 description: anyOf: - type: string - type: 'null' title: Description description: Optional project description examples: - >- This project aligns the model to extract and risk-flag contract clauses. type: object required: - alignment_name - base_model_id - document_ids title: AlignmentProjectCreateRequest description: >- Create an alignment project that fine-tunes a base model on your documents. AlignmentProjectCreateResponse: properties: alignment_id: type: string title: Alignment Id description: Created alignment project identifier examples: - alignment-project-123 status: $ref: '#/components/schemas/ResourceStatus' description: Initial alignment project status default: QUEUED examples: - QUEUED message: anyOf: - type: string - type: 'null' title: Message description: Additional context about the response examples: - Alignment already exists with the same configuration type: object required: - alignment_id title: AlignmentProjectCreateResponse description: Response schema for alignment project creation endpoint ErrorResponse: description: Body of an error response. properties: detail: anyOf: - type: string - $ref: '#/components/schemas/ErrorDetail' description: >- Human readable explanation of the error, or an object with a message and the id of the resource concerned examples: - Model not found title: Detail required: - detail title: ErrorResponse type: object HTTPValidationError: properties: detail: items: $ref: '#/components/schemas/ValidationError' type: array title: Detail type: object title: HTTPValidationError description: Body of a request validation error. ResourceStatus: type: string enum: - QUEUED - PROCESSING - READY - FAILED - STOPPED title: ResourceStatus description: >- Lifecycle status of a resource that runs asynchronously. Stored as is and returned as is, with no translation in between. - `QUEUED`: accepted and waiting for its workflow to start - `PROCESSING`: running - `READY`: finished successfully - `FAILED`: failed at any stage - `STOPPED`: stopped by the user before completion ErrorDetail: additionalProperties: true description: >- Structured error detail: a message, with the id of the resource it concerns as an extra key. properties: message: description: Human readable explanation of the error title: Message type: string required: - message title: ErrorDetail type: object ValidationError: properties: loc: items: anyOf: - type: string - type: integer type: array title: Location msg: type: string title: Message type: type: string title: Error Type type: object required: - loc - msg - type title: ValidationError description: One request validation problem. securitySchemes: ApiKeyAuth: type: http description: Your Nugen API key, sent as a Bearer token scheme: bearer bearerFormat: API key ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.