generated: '2026-07-18'
method: searched
source: https://docs.dottxt.ai/api/overview
docs:
  overview: https://docs.dottxt.ai/api/overview
  authentication: https://docs.dottxt.ai/api/authentication
  chat_completions: https://docs.dottxt.ai/api/chat-completions
  streaming: https://docs.dottxt.ai/json-schema/streaming
  migrate: https://docs.dottxt.ai/migrate-from-other-providers
summary: >-
  dottxt exposes an OpenAI-compatible Chat Completions API. The single most
  important convention is structured output: pass a JSON Schema via
  `response_format` and the output is guaranteed to satisfy it by construction
  (constrained decoding) — no retries, no validation loops.
compatibility:
  standard: openai-chat-completions
  note: >-
    Drop-in for OpenAI SDKs by changing the base URL to
    https://api.dottxt.ai/v1 and the API key.
authentication:
  style: bearer-api-key
  header: 'Authorization: Bearer $DOTTXT_API_KEY'
  key_prefix: sk-dottxt-
structured_output:
  parameter: response_format
  modes:
  - type: json_schema
    description: Output guaranteed to match the supplied JSON Schema.
  - type: json_object
  - type: text
streaming:
  supported: true
  modes:
  - value: false
    description: Single JSON response (default).
  - value: true
    description: Token deltas via Server-Sent Events (SSE).
  - value: patch
    description: >-
      Field-by-field structured streaming as RFC 6902 JSON Patch operations.
      NDJSON by default; set `Accept: text/event-stream` for SSE.
  standards:
  - Server-Sent Events
  - RFC 6902 (JSON Patch)
idempotency:
  supported: false
  note: >-
    No idempotency-key header is documented. Chat completion requests are not
    idempotent; use the `seed` parameter for deterministic sampling.
pagination:
  supported: false
  note: List Models returns the full set in a single `{object:list, data:[]}` envelope.
versioning:
  scheme: uri-path
  current: v1
  base_url: https://api.dottxt.ai/v1
error_envelope:
  shape: openai-error
  example: '{"error": {"code": "", "message": "", "type": "", "param": ""}}'
  cross_reference: errors/dottxt-problem-types.yml
rate_limiting:
  documented: false
  note: >-
    A 429 status ("Rate limit exceeded") is documented as a possible response,
    but no rate-limit headers or published quotas were found in the docs.
cross_links:
  authentication: authentication/dottxt-authentication.yml
  errors: errors/dottxt-problem-types.yml
  lifecycle: lifecycle/dottxt-lifecycle.yml