> ## Documentation Index > Fetch the complete documentation index at: https://docs.compute.prentis.ai/llms.txt > Use this file to discover all available pages before exploring further. # Chat completions > OpenAI-compatible chat completion. Point any OpenAI SDK at this base URL with a Prentis key and nothing else in your code changes. ## Streaming With `stream: true` the response is `text/event-stream`: one `data:` event per chunk, terminated by `data: [DONE]`. The last data event carries `usage` **whether or not** you sent `stream_options.include_usage` — you never have to choose between streaming and knowing what you were billed. If a stream fails partway, chunks already sent are not withdrawn. You get one more event carrying an error object, then `[DONE]`. Treat a stream that ended without `[DONE]` as incomplete. ## Token accounting The `usage` object is the number you are billed on, not an estimate of it. When a stream is cancelled after the model has begun producing output, those tokens are still billed — see [Errors](/errors) for the two codes where that happens. ## OpenAPI ````yaml openapi/prentis.openapi.yaml POST /chat/completions openapi: 3.1.0 info: title: Prentis Inference API version: 0.2.0 summary: One inference API, reachable with an OpenAI client or an Anthropic client. description: > An OpenAI-compatible inference API. Point any OpenAI SDK at this base URL with a Prentis key and nothing else in your code changes. Anthropic clients are served too: the same host answers `/v1/messages`, and `x-api-key` is accepted alongside `Authorization: Bearer`. Parameters this API does not implement are ignored rather than rejected, except the few marked **not supported** below, which return `INVALID_REQUEST` so that a silently wrong answer is never returned in place of an error. servers: - url: https://compute.prentis.ai/v1 description: Production security: - bearerAuth: [] tags: - name: chat - name: completions - name: messages - name: models paths: /chat/completions: post: tags: - chat summary: Create a chat completion description: >- OpenAI-compatible chat completion. Point any OpenAI SDK at this base URL with a Prentis key and nothing else in your code changes. operationId: createChatCompletion parameters: - $ref: '#/components/parameters/IdempotencyKey' requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/ChatCompletionRequest' responses: '200': headers: x-request-id: $ref: '#/components/headers/x-request-id' x-provider-class: $ref: '#/components/headers/x-provider-class' x-config-version: $ref: '#/components/headers/x-config-version' content: application/json: schema: $ref: '#/components/schemas/ChatCompletionResponse' example: id: chatcmpl-01K5S8N4Q2VZ object: chat.completion created: 1789286400 model: deepseek-v4.1-flash choices: - index: 0 message: role: assistant content: >- A distributed store can hold at most two of consistency, availability and partition tolerance at once. Because partitions do happen in practice, the real choice is between answering with stale data and not answering at all. finish_reason: stop usage: prompt_tokens: 28 completion_tokens: 47 total_tokens: 75 text/event-stream: schema: type: string example: >- data: {"id":"chatcmpl-01K5S8N4Q2VZ","object":"chat.completion.chunk","model":"deepseek-v4.1-flash","choices":[{"index":0,"delta":{"role":"assistant","content":"A distributed"}}]} data: {"id":"chatcmpl-01K5S8N4Q2VZ","object":"chat.completion.chunk","model":"deepseek-v4.1-flash","choices":[{"index":0,"delta":{"content":" store can hold"}}]} data: {"id":"chatcmpl-01K5S8N4Q2VZ","object":"chat.completion.chunk","model":"deepseek-v4.1-flash","choices":[{"index":0,"delta":{},"finish_reason":"stop"}],"usage":{"prompt_tokens":28,"completion_tokens":47,"total_tokens":75}} data: [DONE] description: >- OK. With `stream: true` the same call answers `text/event-stream` instead, one chunk per event, terminated by `data: [DONE]`; the last data event carries `usage` whether or not you asked for it. '400': $ref: '#/components/responses/Error400' '401': $ref: '#/components/responses/Error401' '402': $ref: '#/components/responses/Error402' '403': $ref: '#/components/responses/Error403' '404': $ref: '#/components/responses/Error404' '409': $ref: '#/components/responses/Error409' '413': $ref: '#/components/responses/Error413' '429': $ref: '#/components/responses/Error429' '500': $ref: '#/components/responses/Error500' '503': $ref: '#/components/responses/Error503' '504': $ref: '#/components/responses/Error504' x-codeSamples: - lang: curl label: cURL source: |- curl https://compute.prentis.ai/v1/chat/completions \ -H "Authorization: Bearer $PRENTIS_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "deepseek-v4.1-flash", "messages": [ {"role": "system", "content": "You are a concise assistant."}, {"role": "user", "content": "Summarise the CAP theorem in two sentences."} ], "max_tokens": 200, "temperature": 0.7 }' - lang: python label: Python source: >- import os from openai import OpenAI client = OpenAI(base_url="https://compute.prentis.ai/v1", api_key=os.environ["PRENTIS_API_KEY"]) resp = client.chat.completions.create( model="deepseek-v4.1-flash", messages=[ {"role": "system", "content": "You are a concise assistant."}, {"role": "user", "content": "Summarise the CAP theorem in two sentences."}, ], max_tokens=200, temperature=0.7, ) print(resp.choices[0].message.content) print(resp.usage.total_tokens, "tokens billed") - lang: typescript label: TypeScript source: |- import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://compute.prentis.ai/v1", apiKey: process.env.PRENTIS_API_KEY, }); const resp = await client.chat.completions.create({ model: "deepseek-v4.1-flash", messages: [ { role: "system", content: "You are a concise assistant." }, { role: "user", content: "Summarise the CAP theorem in two sentences." }, ], max_tokens: 200, temperature: 0.7, }); console.log(resp.choices[0]?.message.content); components: parameters: IdempotencyKey: name: Idempotency-Key in: header required: false schema: type: string maxLength: 255 description: >- Your own key for making a retry safe (up to 255 characters). It is accepted and echoed back today; de-duplication of replays arrives in a later release, so a retry is currently a second billable call. schemas: ChatCompletionRequest: type: object required: - model - messages additionalProperties: true properties: model: type: string maxLength: 256 description: >- Which model answers the call. A bare slug for platform models (deepseek-v4.1-flash), or a full resource name for something your account owns (accounts/{account}/models/{model}, or a deployment). Never an upstream vendor's own name. messages: type: array minItems: 1 items: $ref: '#/components/schemas/ChatMessage' description: >- The conversation so far, oldest first. Each turn carries a role (system, user, assistant or tool) and its content. stream: type: boolean default: false description: >- Return the answer as server-sent events as it is generated, instead of one response at the end. The last data event carries the usage numbers either way. stream_options: type: object properties: include_usage: type: boolean default: false description: >- Streaming options. include_usage is accepted for compatibility; usage arrives on the final event whether or not you ask for it. max_tokens: type: integer minimum: 1 description: >- Ceiling on how many tokens may be generated. Omit it to use the model's own default. max_completion_tokens: type: integer minimum: 1 description: >- Ceiling on generated tokens. Wins over max_tokens when both are sent. temperature: type: number minimum: 0 maximum: 2 default: 1 description: >- How much randomness to allow when picking each token. Lower is more repeatable, higher is more varied. top_p: type: number minimum: 0 maximum: 1 default: 1 description: >- Nucleus sampling: only consider the most likely tokens up to this share of the probability mass. Tune this or temperature, not both. stop: oneOf: - type: string - type: array items: type: string maxItems: 4 description: >- Up to 4 strings that end generation as soon as the model produces one. The matched string is not included in the output. seed: type: integer description: >- Best effort repeatability: the same seed with the same parameters returns the same answer on models that support it. frequency_penalty: type: number minimum: -2 maximum: 2 default: 0 description: >- Pushes the model away from tokens it has already used often in this answer. Negative values do the opposite. presence_penalty: type: number minimum: -2 maximum: 2 default: 0 description: >- Pushes the model away from tokens that have appeared at all in this answer, which tends to move it on to new ground. user: type: string maxLength: 256 description: >- Your own identifier for the end user behind the call. It is echoed back in this request's log line so you can correlate the two; it is not kept on the usage record. tools: type: array items: $ref: '#/components/schemas/Tool' description: >- Function definitions the model may call. If none of the providers that can serve this model support tools, the call is rejected rather than answered as though you had not asked. tool_choice: oneOf: - type: string enum: - none - auto - required - type: object properties: type: type: string enum: - function function: type: object required: - name properties: name: type: string description: >- Whether the model may, must, or must not call a tool: none, auto, required, or one named function. response_format: type: object properties: type: type: string enum: - text - json_object - json_schema json_schema: type: object description: >- Ask for free text, a JSON object, or JSON matching a schema. Rejected if no provider behind this model can honour it. service_tier: type: string enum: - standard - priority - fast default: standard description: Which service tier serves the call. Only standard is accepted today. 'n': type: integer const: 1 description: |- How many completions to return. **Only `1` is accepted.** Any other value returns `INVALID_REQUEST`. logprobs: type: boolean const: false description: |- Per-token log probabilities. **Only `false` is accepted.** `true` returns `INVALID_REQUEST`. ChatCompletionResponse: type: object required: - id - object - created - model - choices - usage properties: id: type: string object: type: string const: chat.completion created: type: integer model: type: string choices: type: array items: $ref: '#/components/schemas/Choice' usage: $ref: '#/components/schemas/Usage' system_fingerprint: type: string ChatMessage: type: object required: - role properties: role: type: string enum: - system - user - assistant - tool content: oneOf: - type: string - type: array items: $ref: '#/components/schemas/ContentPart' - type: 'null' name: type: string tool_calls: type: array items: $ref: '#/components/schemas/ToolCall' tool_call_id: type: string Tool: type: object required: - type - function properties: type: type: string enum: - function function: type: object required: - name properties: name: type: string parameters: type: object strict: type: boolean Choice: type: object required: - index - message - finish_reason properties: index: type: integer const: 0 message: type: object required: - role properties: role: type: string enum: - assistant content: type: - string - 'null' tool_calls: type: array items: $ref: '#/components/schemas/ToolCall' finish_reason: type: - string - 'null' enum: - stop - length - tool_calls - content_filter - null Usage: type: object required: - prompt_tokens - completion_tokens - total_tokens properties: prompt_tokens: type: integer minimum: 0 completion_tokens: type: integer minimum: 0 total_tokens: type: integer minimum: 0 prompt_tokens_details: type: object properties: cached_tokens: type: integer minimum: 0 media: type: object properties: input_images: type: integer minimum: 0 input_video_seconds: type: number minimum: 0 input_audio_seconds: type: number minimum: 0 output_images: type: integer minimum: 0 output_video_seconds: type: number minimum: 0 output_audio_seconds: type: number minimum: 0 ErrorResponse: type: object required: - error properties: error: type: object required: - message - type - code properties: message: type: string type: type: string enum: - invalid_request_error - authentication_error - permission_error - not_found_error - rate_limit_error - server_error - service_unavailable - billing_error code: type: string enum: - INVALID_API_KEY - TENANT_SUSPENDED - MODEL_NOT_ALLOWED - AGREEMENT_REQUIRED - INVALID_REQUEST - MODEL_NOT_FOUND - MODEL_CAPABILITY_UNSUPPORTED - CONTEXT_LENGTH_EXCEEDED - IDEMPOTENCY_CONFLICT - MEDIA_TOO_LARGE - MEDIA_FETCH_FAILED - MEDIA_UNSUPPORTED - TENANT_RATE_LIMITED - TENANT_BUDGET_EXCEEDED - INSUFFICIENT_BALANCE - PAYMENT_REQUIRED - CONTENT_BLOCKED_INPUT - CONTENT_BLOCKED_OUTPUT - UPSTREAM_RATE_LIMITED - UPSTREAM_UNAVAILABLE - UPSTREAM_TIMEOUT - PARTIAL_RESPONSE_TIMEOUT - UPSTREAM_CONTENT_REJECTED - DEPLOYMENT_SCALING_UP - REQUEST_CANCELLED - REQUEST_DEADLINE_EXCEEDED - INTERNAL - STORAGE_UNAVAILABLE param: type: - string - 'null' ContentPart: oneOf: - type: object required: - type - text properties: type: type: string const: text text: type: string - type: object required: - type - image_url properties: type: type: string const: image_url image_url: type: object required: - url properties: url: type: string detail: type: string enum: - auto - low - high default: auto - type: object required: - type - video_url properties: type: type: string const: video_url video_url: type: object required: - url properties: url: type: string detail: type: string enum: - auto - low - high default: auto max_frames: type: integer minimum: 1 spatial_limit: type: integer minimum: 1 sample_fps: type: number exclusiveMinimum: 0 description: '**Not supported.** Sending it returns `INVALID_REQUEST`.' - type: object required: - type - input_audio properties: type: type: string const: input_audio input_audio: type: object properties: data: type: string url: type: string format: type: string enum: - wav - mp3 - flac - ogg - webm description: '**Not supported.** Sending it returns `INVALID_REQUEST`.' ToolCall: type: object required: - id - type - function properties: id: type: string type: type: string enum: - function function: type: object required: - name - arguments properties: name: type: string arguments: type: string headers: x-request-id: schema: type: string required: true description: >- On every response, successes and errors alike. It is the handle that reaches the whole path your request took -- quote it when you ask us anything about a call. x-provider-class: schema: type: string enum: - primary - fallback description: >- primary or fallback: whether your call was served by the first choice or a standby. It never names who served it. x-config-version: schema: type: string description: >- Which version of the routing configuration was in effect for this call. Useful only when comparing two calls that behaved differently. Retry-After: schema: type: integer minimum: 0 description: How many seconds to wait before retrying, on the responses that set it. responses: Error400: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `CONTENT_BLOCKED_INPUT`, `CONTENT_BLOCKED_OUTPUT`, `CONTEXT_LENGTH_EXCEEDED`, `INVALID_REQUEST`, `MEDIA_FETCH_FAILED`, `MEDIA_UNSUPPORTED`, `MODEL_CAPABILITY_UNSUPPORTED`, `UPSTREAM_CONTENT_REJECTED`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error401: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `INVALID_API_KEY`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error402: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `INSUFFICIENT_BALANCE`, `PAYMENT_REQUIRED`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error403: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `AGREEMENT_REQUIRED`, `MODEL_NOT_ALLOWED`, `TENANT_SUSPENDED`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error404: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `MODEL_NOT_FOUND`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error409: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `IDEMPOTENCY_CONFLICT`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error413: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `MEDIA_TOO_LARGE`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error429: headers: x-request-id: $ref: '#/components/headers/x-request-id' Retry-After: $ref: '#/components/headers/Retry-After' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `TENANT_BUDGET_EXCEEDED`, `TENANT_RATE_LIMITED`, `UPSTREAM_RATE_LIMITED`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error500: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `INTERNAL`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error503: headers: x-request-id: $ref: '#/components/headers/x-request-id' Retry-After: $ref: '#/components/headers/Retry-After' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `DEPLOYMENT_SCALING_UP`, `STORAGE_UNAVAILABLE`, `UPSTREAM_UNAVAILABLE`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. Error504: headers: x-request-id: $ref: '#/components/headers/x-request-id' content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' description: >- `error.code` is one of: `PARTIAL_RESPONSE_TIMEOUT`, `REQUEST_DEADLINE_EXCEEDED`, `UPSTREAM_TIMEOUT`. See the Errors page for what each one means, whether it is safe to retry, and whether it is billed. securitySchemes: bearerAuth: type: http scheme: bearer description: >- `Authorization: Bearer `. On `/v1/messages`, the Anthropic-style `x-api-key: ` header is accepted instead. ````