> ## Documentation Index > Fetch the complete documentation index at: https://docs.usehasp.com/llms.txt > Use this file to discover all available pages before exploring further. # Run HASP's native chat completion for the caller's org > Streams by default, or returns the assembled response. ## OpenAPI ````yaml /openapi/v1.json post /ai/chat openapi: 3.1.0 info: title: HASP AI API version: '2026-07-12' description: >- The HASP Public AI API is the regulated-AI substrate for healthcare and other regulated industries — identity, policy, audit, compliance, and PHI handling for AI inference, exposed via two surfaces: - **Native (`/v1/ai/*`)**: HASP-native chat with full event taxonomy, run lifecycle, and PHI metadata. - **Anthropic-compat (`/v1/messages`)**: Drop-in replacement for `@anthropic-ai/sdk` — change only `baseURL`. All Gateway compliance checks (BAA, credits, PHI policy) apply. All requests require an API key (`Authorization: Bearer hasp_api_live_`). See [Authentication](https://docs.usehasp.com/ai-api/authentication) for key management. servers: - url: https://api.usehasp.com/v1 description: Production security: - http: [] paths: /ai/chat: post: tags: - Chat summary: Run HASP's native chat completion for the caller's org description: Streams by default, or returns the assembled response. operationId: v1.ai.chat requestBody: content: application/json: schema: $ref: '#/components/schemas/ChatRequest' responses: '200': description: '' content: application/json: schema: anyOf: - type: object properties: success: type: boolean data: type: object properties: id: type: string model: {} role: type: string const: assistant content: type: array items: anyOf: - type: object properties: type: type: string const: text text: type: string required: - type - text - type: object properties: type: type: string const: tool_use id: anyOf: - type: string - type: string enum: - '' name: anyOf: - type: string - type: string enum: - '' input: anyOf: - type: string - type: array items: type: string minItems: 0 maxItems: 0 additionalItems: false required: - type - id - name - input stop_reason: type: string usage: anyOf: - type: string - type: array items: type: string minItems: 0 maxItems: 0 additionalItems: false required: - id - model - role - content - stop_reason - usage meta: type: object properties: request_id: type: - string - 'null' required: - request_id required: - success - data - meta - type: string text/event-stream: schema: type: string headers: Transfer-Encoding: required: true schema: type: string enum: - chunked '401': description: Bearer token is missing, malformed, or revoked. content: application/json: schema: type: object example: success: false error: type: authentication code: INVALID_API_KEY message: Bearer token is missing, malformed, or revoked. param: null details: null retryable: false request_id: 01JQREQ7XZQK5N6PZ1VVXHYB8T meta: request_id: 01JQREQ7XZQK5N6PZ1VVXHYB8T '403': $ref: '#/components/responses/AuthorizationException' '404': description: '' content: application/json: schema: anyOf: - type: object properties: success: type: boolean error: type: object properties: type: type: string code: type: string message: type: string const: Conversation not found. param: type: string const: conversation_id details: type: - string - 'null' retryable: type: boolean request_id: type: string required: - type - code - message - param - details - retryable - request_id meta: type: object properties: request_id: type: string required: - request_id errors: type: string required: - success - error - meta - errors - type: object properties: error: type: object properties: message: type: string const: Conversation not found. type: type: string enum: - authentication_error - permission_error - invalid_request_error - not_found_error - rate_limit_error - api_error param: type: string const: conversation_id code: type: string hasp_code: type: string hasp_details: type: string required: - message - type - param - code - hasp_code - hasp_details required: - error - type: object properties: type: type: string const: error error: type: object properties: type: type: string enum: - authentication_error - permission_error - invalid_request_error - not_found_error - rate_limit_error - api_error - overloaded_error message: type: string const: Conversation not found. hasp_code: type: string hasp_request_id: type: string hasp_details: type: string required: - type - message - hasp_code - hasp_request_id - hasp_details required: - type - error '422': $ref: '#/components/responses/ValidationException' '429': description: >- Rate limit exceeded. Retry after the window indicated by `Retry-After`. content: application/json: schema: type: object example: success: false error: type: rate_limited code: RATE_LIMITED message: >- Rate limit exceeded (per-key, per-org, or daily cap, depending on which limiter tripped). param: null details: null retryable: true request_id: 01JQREQ7XZQK5N6PZ1VVXHYB8T meta: request_id: 01JQREQ7XZQK5N6PZ1VVXHYB8T headers: Retry-After: description: Seconds until the rate-limit window resets. schema: type: integer X-RateLimit-Limit: description: The request cap for the current window. schema: type: integer X-RateLimit-Remaining: description: Requests remaining in the current window (0 on a 429). schema: type: integer X-RateLimit-Reset: description: Unix timestamp when the window resets. schema: type: integer '500': description: |- Generic message — matches the streaming path's SSE error event, which never surfaces the raw exception message (may carry upstream/provider details) to the caller. content: application/json: schema: type: object properties: success: type: boolean error: type: object properties: code: type: string const: INFERENCE_UPSTREAM_FAILURE message: type: string const: Inference stream failed. required: - code - message required: - success - error components: schemas: ChatRequest: type: object description: >- Validates POST /v1/ai/chat per PRD §05-public-api. HASP-509: extended from a single-turn `message` field to a full superset — `messages[]` history (same content-block dialect Anthropic-compat uses, so AnthropicMessagesRequestBuilder is reused verbatim for both), `tools`, and the full cross-provider parameter set (`top_p`/`top_k`/`tool_choice`/ `metadata`/`stop_sequences`). `message` (singular) is preserved as shorthand for a single user-turn request — exactly one of `message`/`messages` must be present. properties: message: type: string maxLength: 100000 model: type: - string - 'null' maxLength: 100 conversation_id: type: - string - 'null' description: |- Either a bare ULID (the pre-existing passthrough tag) or a `conv_`-prefixed id naming a server-owned conversation (ADR-AHK8YT). Validated as a pattern rather than `ulid` so the prefixed form is not rejected before the controller can resolve it; which of the two it is decides whether the call is stateful. pattern: ^(conv_)?[0-9A-HJKMNP-TV-Za-hjkmnp-tv-z]{26}$ stream: type: - boolean - 'null' store: type: - boolean - 'null' system: type: - string - 'null' maxLength: 10000 tool_choice: type: - string - 'null' description: |- InferenceRequest's canonical vocabulary directly: 'auto'|'any'| 'none'| — no compat-surface wire shape to normalize. top_p: type: - number - 'null' minimum: 0 maximum: 1 top_k: type: - integer - 'null' minimum: 1 provider_extensions: type: - array - 'null' description: |- Escape hatch for provider-specific wire keys the canonical fields above don't (yet) model — see ADR-ZP99PX. items: type: string stop_sequences: type: - array - 'null' items: type: string metadata: type: object properties: user_id: type: - string - 'null' maxLength: 255 subject: type: object description: |- PRD-phi-reidentification-map §8: contextual subject binding for the audit row fidelity layer. Optional everywhere — when present, both type and id must be set (rule `required_with` on each side). properties: type: type: - string - 'null' maxLength: 64 id: type: - string - 'null' maxLength: 256 on_behalf_of: type: object description: |- Asserted end-user identity: which user of your application took this action. Recorded on the audit chain as an assertion made by your integration — HASP does not verify it, and it is never shown to the model. Distinct from `subject`, which names who or what the action is about. `id` is your stable identifier for the person; `display` is an optional human-readable name shown to audit reviewers. Supported for API-key credentials only. properties: id: type: string maxLength: 256 display: type: - string - 'null' maxLength: 256 messages: type: array items: type: object properties: role: type: string enum: - user - assistant content: type: string description: >- Same content-block dialect as MessagesRequest (Anthropic-compat): plain string, or an array of text/tool_use/tool_result blocks. required: - role - content minItems: 1 tools: type: - array - 'null' description: |- App-domain tool shape — no wire-format translation needed since this is HASP's own surface, not a compat passthrough. items: type: object properties: name: type: string description: type: - string - 'null' input_schema: type: - object - 'null' properties: type: type: - string - 'null' title: ChatRequest responses: AuthorizationException: description: Authorization error content: application/json: schema: type: object properties: message: type: string description: Error overview. required: - message ValidationException: description: Validation error content: application/json: schema: type: object properties: message: type: string description: Errors overview. errors: type: object description: A detailed description of each field that failed validation. additionalProperties: type: array items: type: string required: - message - errors securitySchemes: http: type: http scheme: bearer ````