> ## Documentation Index > Fetch the complete documentation index at: https://docs.goenhance.ai/llms.txt > Use this file to discover all available pages before exploring further. # OC Maker > Two-stage generation in a single request: an LLM turns your prompt into a structured character sheet, then that result is rendered into an image. **You supply the full LLM prompt.** No style presets or prompt templates are applied server-side — compose whatever prompt you need. **Required output contract:** your prompt MUST instruct the model to return a JSON object containing an `oc_prompt` field. That field becomes the image prompt. If the model does not return parseable JSON with `oc_prompt`, the job fails. Optionally include `oc_ratio` in that JSON to let the model pick the aspect ratio; the `ratio` request field overrides it. **Pricing:** 1 token per request (= $0.02). Covers both the text and image stages. Returns an `img_uuid`; poll GET /api/v1/jobs/detail (or use `custom_callback_url`) to get the result. The result contains two entries: the full LLM JSON (`type: "json"`) and the rendered image (`type: "image"`, `image_type: "main_view"`). ## Prompt contract This endpoint runs two stages: a language model turns your prompt into a structured character sheet, then that sheet is rendered into an image. Your prompt **must** instruct the model to return a JSON object containing an `oc_prompt` field — that field becomes the image prompt. If the model returns anything that cannot be parsed as JSON, or the JSON has no `oc_prompt`, the job fails. ```text theme={null} Create a character based on: a stoic knight with silver hair. Return ONLY a JSON object in this format: {"role_name":"","appearance":"","personality":"","oc_prompt":"","oc_ratio":"3:4"} "oc_prompt" must be an English image prompt starting with "Anime style, Single character, full body, front view". ``` Any other keys you ask for (`role_name`, `personality`, `character_backstory`, …) are passed through to the result untouched, so you can shape the character sheet however you like. `oc_ratio` is optional — it lets the model choose the aspect ratio. The `ratio` request field overrides it, and `3:4` is used when neither is present. ## Result Unlike the other image endpoints, the result array contains **two entries**: the full JSON from the text stage, then the rendered image. Poll `GET /api/v1/jobs/detail` or use `custom_callback_url`. ```json theme={null} { "code": 0, "msg": "Success", "data": { "img_uuid": "0f1b8c2e-4a7d-4c31-9f28-6de4a1b9c503", "status": "success", "type": "oc-maker", "start_time": "2026-08-02T09:14:22.108Z", "end_time": "2026-08-02T09:15:04.771Z", "model_id": "-1", "json": [ { "type": "json", "value": { "role_name": "Sir Aldric", "appearance": "Tall knight with long silver hair and a scarred jaw", "personality": "Stoic, loyal, slow to anger", "oc_prompt": "Anime style, Single character, full body, front view, a tall knight with long silver hair...", "oc_ratio": "3:4" }, "duration": 28.4, "link_expired_at": "2026-08-03T09:15:04.771Z" }, { "type": "image", "image_type": "main_view", "value": "https://cdn3.goenhance.ai/user/seedream/9dfdb4b5-7ef7-4fe7-93ff-b9eb76cc431d.jpg", "duration": 28.4, "link_expired_at": "2026-08-03T09:15:04.771Z" } ], "job_type": "oc-maker" } } ``` The `value` of the `type: "json"` entry is an **object**, not a URL string — the only endpoint where this happens. Read the image from the entry with `type: "image"` rather than assuming `json[0]` is the image. A few details worth knowing: * `link_expired_at` applies to the image link. It is also present on the JSON entry, where it has no meaning — ignore it there. * `duration` measures the image stage only, not the text stage, so it under-reports total latency. * Both stages share one `img_uuid` and are billed once. A failure in either stage fails the whole job and refunds the tokens. ## OpenAPI ````yaml POST /api/v1/oc-maker openapi: 3.1.0 info: title: GoEnhance API description: '' version: 1.0.0 servers: - url: https://api.goenhance.ai security: - bearerAuth: [] tags: - name: Basic - name: Video2Video paths: /api/v1/oc-maker: post: tags: - ImageAPI summary: OC Maker description: >- Two-stage generation in a single request: an LLM turns your prompt into a structured character sheet, then that result is rendered into an image. **You supply the full LLM prompt.** No style presets or prompt templates are applied server-side — compose whatever prompt you need. **Required output contract:** your prompt MUST instruct the model to return a JSON object containing an `oc_prompt` field. That field becomes the image prompt. If the model does not return parseable JSON with `oc_prompt`, the job fails. Optionally include `oc_ratio` in that JSON to let the model pick the aspect ratio; the `ratio` request field overrides it. **Pricing:** 1 token per request (= $0.02). Covers both the text and image stages. Returns an `img_uuid`; poll GET /api/v1/jobs/detail (or use `custom_callback_url`) to get the result. The result contains two entries: the full LLM JSON (`type: "json"`) and the rendered image (`type: "image"`, `image_type: "main_view"`). parameters: - name: Authorization in: header description: '' required: false example: '{{Authorization}}' schema: type: string requestBody: content: application/json: schema: type: object properties: prompt: type: string minLength: 1 maxLength: 8000 description: >- Full prompt for the text stage. Must instruct the model to return a JSON object containing an `oc_prompt` field. image_render_model: type: string enum: - seedream - gpt-image-2 description: Rendering backend used for the image stage. system_prompt: type: string maxLength: 4000 description: >- Optional system prompt for the text stage. Defaults to a generic assistant prompt. text_model: type: string maxLength: 100 description: >- Optional LLM model id for the text stage. Defaults to `gemini-3-pro-preview`. ratio: type: string enum: - '1:1' - '2:3' - '3:2' - '4:3' - '3:4' - '16:9' - '9:16' description: >- Optional. Aspect ratio of the generated image. Overrides `oc_ratio` from the LLM output; defaults to 3:4. custom_callback_url: type: string format: uri description: >- Optional. A publicly accessible HTTPS URL. When the task status changes (processing / success / failed), GoEnhance sends a POST request to this URL. The request body is identical to the response of GET /api/v1/jobs/detail. If your server does not respond with HTTP 200, the notification is retried up to 3 times, with a 3-second timeout per attempt. example: https://your-server.com/goenhance/callback required: - prompt - image_render_model example: prompt: A samurai standing in falling cherry blossoms, cinematic ratio: '16:9' examples: '1': summary: Character sheet value: prompt: >- You are an OC generator. Create a character based on: a stoic knight with silver hair. Return ONLY a JSON object: {"role_name":"","appearance":"","personality":"","oc_prompt":"","oc_ratio":"3:4"} The oc_prompt must be an English image prompt starting with "Anime style, Single character, full body, front view". image_render_model: seedream ratio: '3:4' responses: '200': description: '' content: application/json: schema: type: object properties: code: type: integer msg: type: string data: type: object properties: img_uuid: type: string required: - img_uuid required: - code - msg - data examples: '1': summary: Success value: code: 0 msg: Success data: img_uuid: c12b656c-747a-44fd-9c80-add79b0c52d5 '2': summary: Insufficient tokens value: code: 100 msg: tokens is not enough headers: {} '401': description: '' content: application/json: schema: type: object properties: code: type: integer msg: type: string required: - code - msg headers: {} deprecated: false security: [] components: {} ````