openapi: 3.0.1 info: title: Z.AI Agents Paas API description: Z.AI API available endpoints license: name: Z.AI Developer Agreement and Policy url: https://chat.z.ai/legal-agreement/terms-of-service version: 1.0.0 contact: name: Z.AI Developers url: https://chat.z.ai/legal-agreement/privacy-policy email: user_feedback@z.ai servers: - url: https://api.z.ai/api description: Production server security: - bearerAuth: [] tags: - name: Paas paths: /paas/v4/chat/completions: post: description: Create a chat completion model that generates AI replies for given conversation messages. It supports multimodal inputs (text, images, audio, video, file), offers configurable parameters (like temperature, max tokens, tool use), and supports both streaming and non-streaming output modes. parameters: - $ref: '#/components/parameters/AcceptLanguage' requestBody: content: application/json: schema: oneOf: - title: Text Model $ref: '#/components/schemas/ChatCompletionTextRequest' - title: Vision Model $ref: '#/components/schemas/ChatCompletionVisionRequest' examples: Basic Example: value: model: glm-5.1 messages: - role: system content: You are a useful AI assistant. - role: user content: Please tell us about the development of artificial intelligence. temperature: 1 stream: false Stream Example: value: model: glm-5.1 messages: - role: user content: Write a poem about spring. temperature: 1 stream: true Thinking Example: value: model: glm-5.1 messages: - role: user content: Write a poem about spring. thinking: type: enabled stream: true Multi Conversation: value: model: glm-5.1 messages: - role: system content: You are a professional programming assistant. - role: user content: What is recursion? - role: assistant content: Recursion is a programming technique where a function calls itself to solve a problem... What is recursion - role: user content: Can you give me an example of Python recursion? stream: true Image Visual Example: value: model: glm-5v-turbo messages: - role: user content: - type: image_url image_url: url: https://cdn.bigmodel.cn/static/logo/register.png - type: image_url image_url: url: https://cdn.bigmodel.cn/static/logo/api-key.png - type: text text: What are the pics talk about? Video Visual Example: value: model: glm-5v-turbo messages: - role: user content: - type: video_url video_url: url: https://cdn.bigmodel.cn/agent-demos/lark/113123.mov - type: text text: What are the video show about? File Visual Example: value: model: glm-5v-turbo messages: - role: user content: - type: file_url file_url: url: https://cdn.bigmodel.cn/static/demo/demo2.txt - type: file_url file_url: url: https://cdn.bigmodel.cn/static/demo/demo1.pdf - type: text text: What are the files show about? Function Call Example: value: model: glm-5.1 messages: - role: user content: Is there an example of how the weather in Beijing is today? tools: - type: function function: name: get_weather description: Get weather information for the specified city. parameters: type: object properties: city: type: string description: City Name required: - city tool_choice: auto temperature: 0.3 required: true responses: '200': description: Processing successful content: application/json: schema: $ref: '#/components/schemas/ChatCompletionResponse' default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/videos/generations: post: description: 'CogVideoX is a video generation large model developed by Z.AI, equipped with powerful video generation capabilities. Simply inputting text or images allows for effortless video creation. Vidu: A high-performance video large model that combines high consistency and high dynamism, with precise semantic understanding and exceptional reasoning speed.' parameters: - $ref: '#/components/parameters/AcceptLanguage' requestBody: content: application/json: schema: oneOf: - title: CogVideoX-3 $ref: '#/components/schemas/CogVideoX3Request' - title: 'Vidu: Text to Video' $ref: '#/components/schemas/ViduText2VideoRequest' - title: 'Vidu: Image to Video' $ref: '#/components/schemas/ViduImage2VideoRequest' - title: 'Vidu: First & Last Frame to Video' $ref: '#/components/schemas/ViduFrames2VideoRequest' - title: 'Vidu: Ref to Video' $ref: '#/components/schemas/ViduReference2VideoRequest' examples: Text to Video Example: value: model: cogvideox-3 prompt: A cat is playing with a ball. quality: quality with_audio: true size: 1920x1080 fps: 30 Image to Video Example: value: model: cogvideox-3 image_url: https://img.iplaysoft.com/wp-content/uploads/2019/free-images/free_stock_photo.jpg prompt: Make the picture move quality: quality with_audio: true size: 1920x1080 fps: 30 First Last Frame to Video: value: model: cogvideox-3 image_url: - https://gd-hbimg.huaban.com/ccee58d77afe8f5e17a572246b1994f7e027657fe9e6-qD66In_fw1200webp - https://gd-hbimg.huaban.com/cc2601d568a72d18d90b2cc7f1065b16b2d693f7fa3f7-hDAwNq_fw1200webp prompt: Make the picture move quality: quality with_audio: true size: 1920x1080 fps: 30 required: true responses: '200': description: Processing successful. content: application/json: schema: $ref: '#/components/schemas/VideoResponse' default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/async-result/{id}: get: description: This endpoint is used to query the result of an asynchronous request. parameters: - $ref: '#/components/parameters/AcceptLanguage' - name: id in: path required: true schema: type: string description: Task id. responses: '200': description: Processing successful content: application/json: schema: oneOf: - title: Video Generation $ref: '#/components/schemas/AsyncVideoGenerationResponse' - title: Image Generation $ref: '#/components/schemas/AsyncImageGenerationResponse' default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/images/generations: post: summary: Generate Image description: Use [GLM-Image](/guides/image/glm-image) series models to generate high-quality images from text prompts. Through quick and accurate understanding of user text descriptions, `AI` image expression becomes more precise and personalized. requestBody: content: application/json: schema: $ref: '#/components/schemas/CreateImageRequest' examples: Generate Image Example: value: model: glm-image prompt: A cute little kitten sitting on a sunny windowsill, with the background of blue sky and white clouds. size: 1280x1280 required: true responses: '200': description: Processing successful content: application/json: schema: $ref: '#/components/schemas/ImageGenerationResponse' default: description: Request Failed content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/async/images/generations: post: summary: Generate Image (Async) description: Use the [GLM-Image](/guides/image/glm-image) series models to generate high-quality images from text prompts. Through quick and accurate understanding of user text descriptions, `AI` image expression becomes more precise and personalized. Only supports `GLM-Image` model. requestBody: content: application/json: schema: $ref: '#/components/schemas/AsyncCreateImageRequest' examples: Image Generation Example: value: model: glm-image prompt: A cute little kitten sitting on a sunny windowsill, with the background of blue sky and white clouds. size: 1280x1280 required: true responses: '200': description: Processing successful content: application/json: schema: $ref: '#/components/schemas/AsyncResponse' default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/audio/transcriptions: post: summary: Speech to Text description: Use the [GLM-ASR-2512](/guides/audio/glm-asr-2512) model to transcribe audio files into text, supporting multiple languages and real-time streaming transcription. requestBody: content: multipart/form-data: schema: $ref: '#/components/schemas/AudioTranscriptionRequest' example: model: glm-asr-2512 stream: false required: true responses: '200': description: Request processed successfully content: application/json: schema: $ref: '#/components/schemas/AudioTranscriptionResponse' text/event-stream: schema: $ref: '#/components/schemas/AudioTranscriptionStreamResponse' default: description: Request failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/tokenizer: post: summary: Text Tokenizer description: '`Tokenizer` is used to split text into `tokens` recognizable by the model and calculate the count. It receives user input text, processes it through the model for tokenization, and finally returns the corresponding `token` count. It is suitable for text length evaluation, model input estimation, dialogue context truncation, cost calculation, etc.' requestBody: content: application/json: schema: $ref: '#/components/schemas/TokenizerRequest' examples: Text Tokenization Example: value: model: glm-4.6 messages: - role: user content: What opportunities and challenges will the Chinese large model industry face in 2025? required: true responses: '200': description: Business processing successful content: application/json: schema: $ref: '#/components/schemas/TokenizerResponse' default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/layout_parsing: post: summary: Layout Parsing description: Use the [GLM-OCR](/guides/vlm/glm-ocr) model to parse the layout of documents and images and extract text content. Support OCR recognition of images and PDF documents, returning detailed layout information and visualization results. requestBody: content: application/json: schema: $ref: '#/components/schemas/LayoutParsingRequest' examples: Layout Parsing Example: value: model: glm-ocr file: https://cdn.bigmodel.cn/static/logo/introduction.png required: true responses: '200': description: Business processing successful content: application/json: schema: $ref: '#/components/schemas/LayoutParsingResponse' default: description: Request failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/web_search: post: description: The [Web Search](/guides/tools/web-search) is a specialized search engine for large language models. Building upon traditional search engine capabilities like web crawling and ranking, it enhances intent recognition to return results better suited for LLM processing (including webpage titles, URLs, summaries, site names, favicons etc.). parameters: - $ref: '#/components/parameters/AcceptLanguage' requestBody: content: application/json: schema: $ref: '#/components/schemas/WebSearchRequest' required: true responses: '200': description: Processing successful content: application/json: schema: $ref: '#/components/schemas/WebSearchResponse' default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas /paas/v4/files: post: description: This API is designed for uploading auxiliary files (such as glossaries, terminology lists) to support the translation service. It allows users to upload reference materials that can enhance translation accuracy and consistency. requestBody: content: multipart/form-data: schema: type: object properties: purpose: type: string description: Upload purpose (agent) default: agent enum: - agent file: type: string format: binary description: 'File to upload. Limit to `100MB`. Allowed formats: `pdf`, `doc`, `xlsx`, `ppt`, `txt`, `jpg`, `png`.' required: - purpose - file required: true responses: '200': description: Processing successful content: application/json: schema: type: object properties: id: type: string description: Unique identifier of the uploaded file. object: type: string description: Object type. bytes: type: integer description: File size in bytes. filename: type: string description: Name of the uploaded file. purpose: type: string description: Purpose of the uploaded file. created_at: type: integer description: Timestamp of file creation. default: description: The request has failed. content: application/json: schema: $ref: '#/components/schemas/Error' tags: - Paas components: schemas: ViduText2VideoRequest: allOf: - type: object properties: model: type: string description: The model code to be called. enum: - viduq1-text prompt: type: string description: Text description of the video, maximum input length of 512 characters. style: type: string description: 'Style Default: `general` Optional values: `general` , `anime` - `general`: General style, can be controlled using prompts to define the style. - `anime`: Anime style, optimized for anime-specific visuals. The style can be controlled using different anime-themed prompts.' enum: - general - anime duration: type: integer description: 'Video duration parameter. Default: `5` , Optional: `5`.' example: 5 enum: - 5 aspect_ratio: type: string description: 'Aspect ratio Default: `16:9`, Optional values: `16:9`, `9:16`, `1:1`' example: '16:9' enum: - '16:9' - '9:16' - '1:1' size: type: string description: 'Resolution parameter Default: `1920x1080`, Optional: `1920x1080`' example: 1920x1080 enum: - 1920x1080 movement_amplitude: type: string description: 'Motion amplitude Default: `auto` , Optional values: `auto` ,`small` ,`medium` ,`large`' example: auto enum: - auto - small - medium - large required: - model - prompt - $ref: '#/components/schemas/VideoCommonRequest' WebSearchToolSchema: type: object title: Web Search properties: type: type: string default: web_search enum: - web_search web_search: $ref: '#/components/schemas/WebSearchObject' required: - type - web_search additionalProperties: false AudioTranscriptionResponse: type: object properties: id: type: string description: Task ID created: type: integer format: int64 description: Request creation time, as a `Unix` timestamp in seconds. request_id: type: string description: Passed by the client, must be unique. A unique identifier to distinguish each request. If not provided by the client, the platform will generate one by default. model: type: string description: Model name text: type: string description: The complete transcribed content of the audio. RetrievalToolSchema: type: object title: Retrieval properties: type: type: string default: retrieval enum: - retrieval retrieval: $ref: '#/components/schemas/RetrievalObject' required: - type - retrieval additionalProperties: false Error: required: - code - message type: object description: The request has failed. properties: code: type: integer format: int32 description: Error code. message: type: string description: Error message. AsyncVideoGenerationResponse: type: object properties: model: type: string description: Model name. task_status: type: string description: 'Processing status, `PROCESSING` (processing), `SUCCESS` (success), `FAIL` (failure). Note: Processing status needs to be obtained via query.' video_result: type: array description: Video generation results. items: type: object properties: url: type: string description: Video URL. cover_image_url: type: string description: Video cover URL. request_id: type: string description: Task number submitted by the user during the client request or generated by the platform. RetrievalObject: type: object properties: knowledge_id: type: string description: Knowledge base ID, created or obtained from the platform prompt_template: type: string description: 'Prompt template for requesting the model, a custom request template containing placeholders `{{ knowledge }}` and `{{ question }}`. Default template: Search for the answer to the question `{{question}}` in the document `{{ knowledge }}`. If an answer is found, respond only using statements from the document; if no answer is found, use your own knowledge to answer and inform the user that the information is not from the document. Do not repeat the question, start the answer directly.' required: - knowledge_id LayoutDetail: type: object description: Layout detail element properties: index: type: integer description: Element index example: 1 label: type: string description: 'Element type: image for images, text for text content, formula for inline formulas, table for tables' enum: - image - text - formula - table example: text bbox_2d: type: array description: Normalized element coordinates [x1,y1,x2,y2] items: type: number minimum: 0 maximum: 1 minItems: 4 maxItems: 4 example: - 0.1 - 0.1 - 0.5 - 0.3 content: type: string description: Element content (text / image URL / table HTML) example: This is the content of the element height: type: integer description: Page height example: 800 width: type: integer description: Page width example: 600 required: - index - label DataInfo: type: object description: Document basic information properties: num_pages: type: integer description: Total number of document pages example: 5 pages: type: array description: Document page count information items: $ref: '#/components/schemas/PageInfo' required: - num_pages TokenizerRequest: type: object required: - model - messages properties: model: type: string description: The model code to be called. example: glm-4.6 default: glm-4.6 enum: - glm-4.6 - glm-4.6v - glm-4.5 messages: type: array description: 'The current conversation message list as the model’s prompt input, provided in JSON array format, e.g.,`{“role”: “user”, “content”: “Hello”}`. Possible message types include system messages, user messages. Note: The input must not consist of system or assistant messages only.' items: oneOf: - title: User Message type: object properties: role: type: string enum: - user description: Role of the message author default: user content: oneOf: - type: array description: Multimodal message content, supports text, images, video, file items: $ref: '#/components/schemas/VisionMultimodalContentItem' - type: string description: Text message content (can switch to multimodal message above) example: What opportunities and challenges will the Chinese large model industry face in 2025? required: - role - content - title: System Message type: object properties: role: type: string enum: - system description: Role of the message author default: system content: oneOf: - type: string description: Message text content example: You are a helpful assistant. required: - role - content - title: Assistant Message type: object description: Can include tool calls properties: role: type: string enum: - assistant description: Role of the message author default: assistant content: oneOf: - type: string description: Text message content example: I'll help you with that analysis. required: - role minItems: 1 tools: type: array description: List of tools the model can call. Supports up to `128` functions. anyOf: - items: $ref: '#/components/schemas/FunctionToolSchema' request_id: type: string description: Passed by the client, must be unique. If empty, it will be generated by default. user_id: type: string description: Unique `ID` of the end user ImageGenerationResponse: type: object properties: created: type: integer example: 1760335349 description: Request creation time, in `Unix` timestamp format, unit is seconds. data: type: array description: Array, containing the generated image `URL`. Currently, the array only contains one image. items: type: object properties: url: type: string description: Image link. The temporary link expires after `30` days, please store it promptly. required: - url content_filter: type: array description: Array, containing content safety related information. items: type: object properties: role: type: string description: Safety enforcement stage, including `role = assistant` model inference, `role = user` user input, `role = history` historical context. enum: - assistant - user - history level: type: integer description: Severity level `level 0-3`, `level 0` is most severe, `3` is least severe. minimum: 0 maximum: 3 CogVideoX3Request: allOf: - type: object properties: model: type: string description: The model code to be called. enum: - cogvideox-3 prompt: type: string description: Text description of the video, maximum input length of 512 characters. Either image_url or prompt must be provided, or both. quality: type: string description: "Output mode, default is `speed`.\n- `quality`: Prioritizes quality, higher generation quality. \n- `speed`: Prioritizes speed, faster generation time, relatively lower quality." example: speed enum: - speed - quality with_audio: type: boolean description: 'Whether to generate AI sound effects. Default: `false` (no sound effects).' example: false image_url: type: array description: 'Provide an image based on which content will be generated. If this parameter is passed, the system will operate based on this image. Supports passing images via URL or Base64 encoding. Image requirements: images support `.png`, `.jpeg`, `.jpg` formats; image size: no more than `5M`. Either image_url and prompt can be used, or both can be passed simultaneously. First and last frames: supports inputting two images. The first uploaded image is regarded as the first frame, and the second image is regarded as the last frame. The model will generate the video based on the images passed in this parameter. First and last frame mode only supports `speed` mode' items: oneOf: - title: Image URL type: string format: uri example: https://example.com/image.png - title: Base64 Encoded Image type: string format: byte example: data:image/png;base64, XXX size: type: string description: 'Default value: if not specified, the short side of the generated video is 1080 by default, and the long side is determined according to the original image ratio. Maximum support for 4K resolution. Resolution options: "1280x720", "720x1280", "1024x1024", "1080x1920", "2048x1080", "3840x2160"' example: 1920x1080 enum: - 1280x720 - 720x1280 - 1024x1024 - 1920x1080 - 1080x1920 - 2048x1080 - 3840x2160 fps: type: integer description: 'Video frame rate (FPS), optional values are `30` or `60`. Default: `30`.' example: 30 enum: - 30 - 60 duration: type: integer description: Video duration, default is 5 seconds, supports `5` and `10` seconds. example: 5 enum: - 5 - 10 required: - model - $ref: '#/components/schemas/VideoCommonRequest' ViduReference2VideoRequest: allOf: - type: object properties: model: type: string description: The model code to be called. enum: - vidu2-reference prompt: type: string description: Text description of the video, maximum input length of 512 characters. Either image_url or prompt must be provided, or both. image_url: type: array description: 'Image reference Supports input of 1 to 3 images. The model will use the themes from the images provided in this parameter as references to generate a video with consistent subjects. 1. Supports image URLs or images encoded in Base64 (ensure accessibility; it is recommended to prioritize using image URLs). 2. Supported formats: `png`, `jpeg`, `.jpg`, `webp`. 3. Image resolution must not be smaller than `128x128`, and the aspect ratio must be less than `1:4` or `4:1`. 4. Image file size must not exceed `50 MB`. 5. Note: After Base64 decoding, the byte length must be less than 50 MB, and the encoding must include the proper content-type string, such as `data:image/png;base64,{base64_encode}`.' items: type: string minLength: 1 minItems: 1 maxItems: 3 duration: title: vidu2-reference type: integer description: 'Video duration parameter. Default: `4` , Optional: `4`.' example: 4 enum: - 4 aspect_ratio: type: string description: 'Aspect ratio Default: `16:9`, Optional values: `16:9`, `9:16`, `1:1`' example: '16:9' enum: - '16:9' - '9:16' - '1:1' size: title: 'vidu2-reference ' type: string description: 'Resolution parameter Default: `1280x720`, Optional: `1280x720`' example: 1280x720 enum: - 1280x720 movement_amplitude: type: string description: 'Motion amplitude Default: `auto` , Optional values: `auto` ,`small` ,`medium` ,`large`' example: auto enum: - auto - small - medium - large with_audio: type: boolean description: Add background music to the generated video. required: - model - $ref: '#/components/schemas/VideoCommonRequest' WebSearchRequest: type: object properties: search_engine: type: string description: "The search engine code to call.\n search-prime: Z.AI Premium Version Search Engine" example: search-prime default: search-prime enum: - search-prime search_query: type: string description: The content to be searched. count: type: integer description: "The number of results to return\nFillable range: `1-50`, maximum `50` results per single search\nDefault is `10`\nSupported search engines: \n`search_pro_jina`." minimum: 1 maximum: 50 search_domain_filter: type: string description: "Used to limit the scope of search results and only return content from specified whitelist domains.\nWhitelist: Directly enter the domain name (e.g., `www.example.com`)\nSupported search engines: \n`search_pro_jina`" search_recency_filter: type: string description: "Search for webpages within a specified time range.\nDefault is `noLimit`\nFillable values:\n`oneDay`: within one day\n`oneWeek`: within one week\n`oneMonth`: within one month\n`oneYear`: within one year\n`noLimit`: no limit (default)\nSupported search engines: \n`search_pro_jina`" enum: - oneDay - oneWeek - oneMonth - oneYear - noLimit request_id: type: string description: User-provided unique identifier for distinguishing requests. If not provided, the platform will generate one. user_id: type: string description: 'Unique ID of the end user, helping the platform intervene in illegal activities, inappropriate content generation, or other abuses. ID length: 6 to 128 characters.' required: - search_engine - search_query VideoResponse: type: object properties: model: description: Model name used in this call. type: string id: description: Task order number generated by the Z.AI, use this order number when calling the request result interface. type: string request_id: description: Task number submitted by the user during the client request or generated by the platform. type: string task_status: description: Processing status, `PROCESSING (processing)`,` SUCCESS (success)`, `FAIL (failure)`. Results need to be obtained via query. type: string AsyncCreateImageRequest: type: object required: - model - prompt properties: model: type: string description: Model code enum: - glm-image example: glm-image prompt: type: string description: The text description of the image to be generated. example: A cute little kitten. quality: type: string description: 'The quality of the generated image. `hd`: Generates a more detailed and rich image with higher overall consistency, takes about `20` seconds.' enum: - hd default: hd size: type: string description: 'Image size, recommended enum values: `1280x1280` (default), `1568x1056`, `1056x1568`, `1472x1088`, `1088x1472`, `1728x960`, `960x1728`. Custom parameter: Both width and height must be between `1024px-2048px`, and must be divisible by `32`, and the maximum pixel count must not exceed `2^22px`.' default: 1280x1280 example: 1280x1280 user_id: type: string description: 'Unique ID of the end user, helping the platform intervene in illegal activities, inappropriate content generation, or other abuses. ID length: 6 to 128 characters.' minLength: 6 maxLength: 128 LayoutParsingRequest: type: object required: - model - file properties: model: type: string description: 'Model code: `glm-ocr`' example: glm-ocr enum: - glm-ocr file: type: string description: 'Image or PDF document to be recognized, supports URL and base64. Supported image formats: PDF, JPG, PNG. Single image ≤10MB, PDF ≤50MB, maximum support 100 pages' example: https://cdn.bigmodel.cn/static/logo/introduction.png return_crop_images: type: boolean description: Whether to return screenshot information default: false need_layout_visualization: type: boolean description: Whether to return detailed layout image result information default: false start_page_id: type: integer description: Start page number for parsing when PDF is provided minimum: 1 end_page_id: type: integer description: End page number for parsing when PDF is provided minimum: 1 request_id: type: string description: Unique request identifier, automatically generated if not provided example: req_123456789 user_id: type: string description: 'End user ID for abuse monitoring. Length: 6-128 characters' minLength: 6 maxLength: 128 example: user_123456 VisionMultimodalContentItem: oneOf: - title: Text type: object properties: type: type: string enum: - text description: Content type is text default: text text: type: string description: Text content required: - type - text additionalProperties: false - title: Image type: object properties: type: type: string enum: - image_url description: Content type is image URL default: image_url image_url: type: object description: Image information properties: url: type: string description: Image URL or Base64 encoding. Image size limit is under 5M per image, with pixels not exceeding 6000*6000. GLM-5V GLM4.6V series are limited to 150 sheets, GLM4.5V limit 50 sheets. Supports jpg, png, jpeg formats. required: - url additionalProperties: false required: - type - image_url additionalProperties: false - title: Video type: object properties: type: type: string enum: - video_url description: Content type is video URL default: video_url video_url: type: object description: Video information. properties: url: type: string description: Video URL address.The video size is limited to within 200 MB, GLM-5V GLM4.6V series are limited to 2 videos, GLM4.5V limit 1 video, and the format supports `mp4`，`mkv`，`mov`. required: - url additionalProperties: false required: - type - video_url additionalProperties: false - title: File type: object properties: type: type: string enum: - file_url description: Content type is file URL, not support passing both the `file_url` and `image_url` or `video_url` parameters at the same time. default: file_url file_url: type: object description: File information. properties: url: type: string description: File URL address. Only GLM-5V-Turbo, GLM-4.6V, GLM-4.5V supported. Supports formats such as pdf、txt、word、jsonl、xlsx、pptx, with a maximum of 50. required: - url additionalProperties: false required: - type - file_url additionalProperties: false ChatCompletionResponseMessage: type: object properties: role: type: string description: Current conversation role, default is ‘assistant’ (model) example: assistant content: type: string description: "Current conversation content. Hits function is null, otherwise returns model inference result. \nFor the GLM-4.5V series models, the output may contain the reasoning process tags ` ` or the text boundary tags `<|begin_of_box|> <|end_of_box|>`." reasoning_content: type: string description: Reasoning content, supports by GLM-4.5 series. tool_calls: type: array description: Function names and parameters generated by the model that should be called. items: $ref: '#/components/schemas/ChatCompletionResponseMessageToolCall' WebSearchObject: type: object properties: enable: type: boolean description: 'Whether to enable search functionality. Default is `false`. Set to true to `enable`.' search_engine: type: string description: 'Type of search engine. Default is `search_pro_jina`. Supports: `search_pro_jina`.' enum: - search_pro_jina search_query: type: string description: Force trigger a search count: type: integer description: 'Number of returned results Range: `1-50`, max `50` results per search Default is `10` Supported engines: `search_pro_jina` ' minimum: 1 maximum: 50 search_domain_filter: type: string description: 'Limits search results to specified whitelisted domains. Whitelist: input domains directly (e.g., www.example.com) Supported engines: `search_pro_jina`' search_recency_filter: type: string description: 'Limits search to a specific time range. Default is `noLimit` Values: `oneDay`, within a day `oneWeek`, within a week `oneMonth`, within a month `oneYear`, within a year `noLimit`, no limit (default) Supported engines: `search_pro_jina`' enum: - oneDay - oneWeek - oneMonth - oneYear - noLimit content_size: type: string description: 'Number of characters for webpage summaries. Default is `medium` `medium`: Balanced mode for most queries. 400-600 characters `high`: Maximizes context for comprehensive answers, 2500 characters.' enum: - medium - high result_sequence: type: string description: 'Specifies whether search results are shown before or after model response. Options: `before`, `after`. Default is `after`' enum: - before - after search_result: type: boolean description: 'Whether to return search results in the response. Default is `false`' require_search: type: boolean description: 'Whether to force model response based on search result. Default is `false`' search_prompt: type: string description: 'Prompt to customize how search results are processed. Default Prompt: `You are an intelligent Q&A expert with the ability to synthesize information, recognize time, understand semantics, and clean contradictory data. The current date is {{current_date}}. Use this as the only time reference. Based on the following information, provide a comprehensive and accurate answer to the user''s question.Only extract valuable content for the answer. Ensure the answer is timely and authoritative. State the answer directly without citing data sources or internal processes.`' required: - search_engine AudioTranscriptionRequest: type: object required: - file - model properties: file: type: string format: binary description: 'The audio file to be transcribed. Supported audio file formats: `.wav / .mp3`. Specifications: file size ≤ `25 MB`, audio duration ≤ `30 seconds`.' file_base64: type: string description: Base64 encoded audio file. Only one of file_base64 or file needs to be provided (if both are provided, file takes precedence). model: type: string description: The model ID to invoke. enum: - glm-asr-2512 default: glm-asr-2512 prompt: type: string description: In long text scenarios, you can provide previous transcription results as context. Recommended to be less than 8000 characters. hotwords: type: array description: 'Hotword list to improve recognition accuracy for domain-specific vocabulary. Format example: ["person_name","place_name"]. Recommended not to exceed 100 items.' items: type: string maxItems: 100 stream: type: boolean default: false description: 'This parameter should be set to `false` or omitted when using synchronous calls. It indicates that the model returns all content at once after generating all content. Default is `false`. If set to `true`, the model will return generated content in chunks via standard `Event Stream`. When the `Event Stream` ends, a `data: [DONE]` message will be returned.' request_id: type: string description: Passed by the client, must be unique. A unique identifier to distinguish each request. If not provided by the client, the platform will generate one by default. user_id: type: string description: 'A unique `ID` for the end user, helping the platform intervene in illegal activities, generation of illegal or inappropriate content, or other abusive behaviors by end users. `ID` length requirement: at least `6` characters, at most `128` characters.' LayoutParsingResponse: type: object properties: id: type: string description: Task ID example: task_123456789 created: type: integer format: int64 description: Request creation time, Unix timestamp in seconds example: 1727156815 model: type: string description: Model name example: GLM-OCR md_results: type: string description: Recognition result in Markdown format example: '# Doc title This is the document content...' layout_details: type: array description: Detailed layout information items: type: array items: $ref: '#/components/schemas/LayoutDetail' layout_visualization: type: array description: Recognition result image URLs items: type: string data_info: $ref: '#/components/schemas/DataInfo' usage: type: object description: Token usage statistics returned when the model call ends. properties: prompt_tokens: type: number description: Number of tokens in user input completion_tokens: type: number description: Number of output tokens prompt_tokens_details: type: object properties: cached_tokens: type: number description: Number of tokens served from cache total_tokens: type: integer description: Total number of tokens request_id: type: string description: Request ID example: req_123456789 required: - id - created - model ViduImage2VideoRequest: allOf: - type: object properties: model: type: string description: The model code to be called. enum: - viduq1-image - vidu2-image prompt: type: string description: Text description of the video, maximum input length of 512 characters. Either image_url or prompt must be provided, or both. image_url: type: string description: 'The model will use the image provided in this parameter as the first frame to generate the video. Only `1` image is supported. Supported formats: `png` , `jpeg` , `jpg` , `webp` . Image aspect ratio must be less than `1:4` or `4:1`. Image file size must not exceed `50MB`. Note: After Base64 decoding, the byte length must be less than 50 MB, and the encoding must include the appropriate content type string (e.g., `data:image/png;base64,{base64_encode}`).' oneOf: - title: Image URL type: string format: uri example: https://example.com/image.png - title: Base64 Encoded Image type: string format: byte example: data:image/png;base64, XXX duration: oneOf: - title: viduq1-image type: integer description: 'Video duration parameter. Default: `5` , Optional: `5`.' example: 5 enum: - 5 - title: viduq2-image type: integer description: 'Video duration parameter. Default: `4` , Optional: `4`.' example: 4 enum: - 4 size: oneOf: - title: viduq1-image type: string description: 'Resolution parameter Default: `1920x1080`, Optional: `1920x1080`' example: 1920x1080 enum: - 1920x1080 - title: viduq2-image type: string description: 'Resolution parameter Default: `1280x720`, Optional: `1280x720`' example: 1280x720 default: 1280x720 enum: - 1280x720 movement_amplitude: type: string description: 'Motion amplitude Default: `auto` , Optional values: `auto` ,`small` ,`medium` ,`large`' example: auto enum: - auto - small - medium - large with_audio: type: boolean description: Add background music to the generated video. required: - model - $ref: '#/components/schemas/VideoCommonRequest' TokenizerResponse: type: object properties: created: type: integer format: int64 example: 1727156815 id: type: string example: 20241120141244890ab4ee4af84acf description: The task sequence number generated by the Zhipu AI Open Platform. Please use this number when calling the request result interface. request_id: type: string example: '1' description: The task number submitted by the client or generated by the platform when the request was initiated. usage: type: object properties: prompt_tokens: type: number description: Prompt tokens in this input image_tokens: type: number description: Image tokens in this input video_tokens: type: number description: Video tokens in this input total_tokens: type: number description: Total tokens in this input required: - id - usage FunctionParameters: type: object description: Parameters defined using JSON Schema. Must pass a JSON Schema object to accurately define accepted parameters. Omit if no parameters are needed when calling the function. additionalProperties: true WebSearchResponse: type: object properties: id: description: Task ID. type: string created: description: Request creation time, Unix timestamp in seconds. type: integer search_result: description: Search results. type: array items: $ref: '#/components/schemas/WebSearchObjectResponse' AsyncResponse: type: object properties: model: description: Model name used in this call. type: string id: description: Task order number generated by the platform, use this order number when calling the async result interface. type: string request_id: description: Task number submitted by the user during the client request or generated by the platform. type: string task_status: description: Processing status, `PROCESSING (processing)`, `SUCCESS (success)`, `FAIL (failure)`. Results need to be obtained via query. type: string AudioTranscriptionStreamResponse: type: object properties: id: type: string description: Task ID created: type: integer format: int64 description: Request creation time, as a `Unix` timestamp in seconds. model: type: string description: Model name type: type: string description: Audio transcription event type. `transcript.text.delta` indicates transcription in progress, `transcript.text.done` indicates transcription completed. delta: type: string description: Incremental audio transcription information returned by the model. FunctionToolSchema: type: object title: Function Call properties: type: type: string default: function enum: - function function: $ref: '#/components/schemas/FunctionObject' required: - type - function additionalProperties: false CreateImageRequest: type: object required: - model - prompt properties: model: type: string description: Model code enum: - glm-image - cogview-4-250304 example: glm-image prompt: type: string description: The text description of the image to be generated. example: A cute little kitten. quality: type: string description: 'The quality of the generated image. `glm-image` default is `hd`, others model is `standard`. `hd`: Generates a more detailed and rich image with higher overall consistency, but takes about `20` seconds. `standard`: Generates an image quickly, suitable for scenarios with higher requirements for generation speed, takes about `5-10` seconds.' enum: - hd - standard default: hd size: type: string description: "Image size. `glm-image` recommended enum values: `1280x1280` (default), `1568x1056`, `1056x1568`, `1472x1088`, `1088x1472`, `1728x960`, `960x1728`. Custom parameter: Both width and height must be between `1024px-2048px`, and must be divisible by `32`, and the maximum pixel count must not exceed `2^22px`. \nOthers model recommended enum values: `1024x1024` (default), `768x1344`, `864x1152`, `1344x768`, `1152x864`, `1440x720`, `720x1440`. Custom parameter: Both width and height must be between `512px-2048px`, and must be divisible by `16`, and the maximum pixel count must not exceed `2^21px`." default: 1280x1280 example: 1280x1280 user_id: type: string description: 'Unique ID of the end user, helping the platform intervene in illegal activities, inappropriate content generation, or other abuses. ID length: 6 to 128 characters.' minLength: 6 maxLength: 128 ChatCompletionVisionRequest: required: - model - messages type: object properties: model: type: string description: The model code to be called. GLM-5V-Turbo are the new generation of visual reasoning models. `AutoGLM-Phone-Multilingual` is mobile intelligent assistant model. example: glm-5v-turbo default: glm-5v-turbo enum: - glm-5v-turbo - glm-4.6v - autoglm-phone-multilingual - glm-4.6v-flash - glm-4.6v-flashx - glm-4.5v messages: type: array description: 'The current conversation message list as the model’s prompt input, provided in JSON array format, e.g.,`{“role”: “user”, “content”: “Hello”}`. Possible message types include system messages, user messages. Note: The input must not consist of system or assistant messages only.' items: oneOf: - title: User Message type: object properties: role: type: string enum: - user description: Role of the message author default: user content: oneOf: - type: array description: Multimodal message content, supports text, images, video, file items: $ref: '#/components/schemas/VisionMultimodalContentItem' - type: string description: Text message content (can switch to multimodal message above) example: What opportunities and challenges will the Chinese large model industry face in 2025? required: - role - content - title: System Message type: object properties: role: type: string enum: - system description: Role of the message author default: system content: oneOf: - type: string description: Message text content example: You are a helpful assistant. required: - role - content - title: Assistant Message type: object description: Can include tool calls properties: role: type: string enum: - assistant description: Role of the message author default: assistant content: oneOf: - type: string description: Text message content example: I'll help you with that analysis. required: - role minItems: 1 request_id: type: string description: Passed by the user side, needs to be unique; used to distinguish each request. If not provided by the user side, the platform will generate one by default. do_sample: type: boolean example: true default: true description: When do_sample is true, sampling strategy is enabled; when do_sample is false, sampling strategy parameters such as temperature and top_p will not take effect. Default value is `true`. stream: type: boolean example: false default: false description: 'This parameter should be set to false or omitted when using synchronous call. It indicates that the model returns all content at once after generating all content. Default value is false. If set to true, the model will return the generated content in chunks via standard Event Stream. When the Event Stream ends, a `data: [DONE]` message will be returned.' thinking: $ref: '#/components/schemas/ChatThinking' temperature: type: number description: 'Sampling temperature, controls the randomness of the output, must be a positive number within the range: `[0.0, 1.0]`. The GLM-5V-Turbo, GLM-4.6V, GLM-4.5V series default value is `0.8`, the autoglm-phone-multilingual default value is `0.0`.' format: float example: 0.8 default: 0.8 minimum: 0 maximum: 1 top_p: type: number description: 'Another method of temperature sampling, value range is: `[0.01, 1.0]`, value range is: `[0.01, 1.0]`. The GLM-5V-Turbo, GLM-4.6V, GLM-4.5V series default value is `0.6`, the autoglm-phone-multilingual default value is `0.85`.' format: float example: 0.6 default: 0.6 minimum: 0.01 maximum: 1 max_tokens: type: integer description: The maximum number of tokens for model output, the GLM-5V-Turbo supports 128K maximum output, GLM-4.6V series supports 32K maximum output, the GLM-4.5V series supports 16K maximum output, the autoglm-phone-multilingual supports 4K maximum output. example: 1024 minimum: 1 maximum: 131072 tools: type: array description: 'A list of tools the model may call. Only support by GLM-4.6V series and autoglm-phone-multilingual. Use this to provide a list of functions the model may generate JSON inputs for. A max of 128 functions are supported. ' items: anyOf: - $ref: '#/components/schemas/FunctionToolSchema' tool_choice: oneOf: - type: string enum: - auto description: Used to control how the model selects which function to call. This is only applicable when the tool type is function. The default value is auto, and only auto is supported. description: Controls how the model selects a tool. stop: type: array description: Stop word list. Generation stops when the model encounters any specified string. Currently, only one stop word is supported, in the format ["stop_word1"]. items: type: string maxItems: 1 user_id: type: string description: Unique ID for the end user, 6–128 characters. Avoid using sensitive information. minLength: 6 maxLength: 128 ChatCompletionTextRequest: required: - model - messages type: object properties: model: type: string description: The model code to be called. GLM-5.1, GLM-5, GLM-5-Turbo are the latest flagship model series, foundational models specifically designed for agent applications. example: glm-5.1 default: glm-5.1 enum: - glm-5.1 - glm-5-turbo - glm-5 - glm-4.7 - glm-4.7-flash - glm-4.7-flashx - glm-4.6 - glm-4.5 - glm-4.5-air - glm-4.5-x - glm-4.5-airx - glm-4.5-flash - glm-4-32b-0414-128k messages: type: array description: 'The current conversation message list as the model’s prompt input, provided in JSON array format, e.g.,`{“role”: “user”, “content”: “Hello”}`. Possible message types include system messages, user messages, assistant messages, and tool messages. Note: The input must not consist of system messages or assistant messages only.' items: oneOf: - title: User Message type: object properties: role: type: string enum: - user description: Role of the message author default: user content: oneOf: - type: string description: Text message content example: What opportunities and challenges will the Chinese large model industry face in 2025? required: - role - content - title: System Message type: object properties: role: type: string enum: - system description: Role of the message author default: system content: oneOf: - type: string description: Message text content example: You are a helpful assistant. required: - role - content - title: Assistant Message type: object description: Can include tool calls properties: role: type: string enum: - assistant description: Role of the message author default: assistant content: oneOf: - type: string description: Text message content example: I'll help you with that analysis. tool_calls: type: array description: Tool call messages generated by the model. When this field is provided, content is usually empty. items: type: object properties: id: type: string description: Tool call ID type: type: string description: Tool type, supports web_search, retrieval, function enum: - function - web_search - retrieval function: type: object description: Function call information, not empty when type is function properties: name: type: string description: Function name arguments: type: string description: Function parameters, JSON format string required: - name - arguments required: - id - type required: - role - title: Tool Message type: object properties: role: type: string enum: - tool description: Role of the message author default: tool content: oneOf: - type: string description: Message text content example: 'Function executed successfully with result: ...' tool_call_id: type: string description: Indicates the tool call ID corresponding to this message required: - role - content - tool_call_id minItems: 1 request_id: type: string description: Passed by the user side, needs to be unique; used to distinguish each request. If not provided by the user side, the platform will generate one by default. do_sample: type: boolean example: true default: true description: When do_sample is true, sampling strategy is enabled; when do_sample is false, sampling strategy parameters such as temperature and top_p will not take effect. Default value is `true`. stream: type: boolean example: false default: false description: 'This parameter should be set to false or omitted when using synchronous call. It indicates that the model returns all content at once after generating all content. Default value is false. If set to true, the model will return the generated content in chunks via standard Event Stream. When the Event Stream ends, a `data: [DONE]` message will be returned.' thinking: $ref: '#/components/schemas/ChatThinking' temperature: type: number description: 'Sampling temperature, controls the randomness of the output, must be a positive number within the range: `[0.0, 1.0]`. The GLM-5.1, GLM-5, GLM-4.7, GLM-4.6 series default value is `1.0`, GLM-4.5 series default value is `0.6`, GLM-4-32B-0414-128K default value is `0.75`.' format: float example: 1 default: 1 minimum: 0 maximum: 1 top_p: type: number description: 'Another method of temperature sampling, value range is: `[0.01, 1.0]`. The GLM-5.1, GLM-5, GLM-4.7, GLM-4.6, GLM-4.5 series default value is `0.95`, GLM-4-32B-0414-128K default value is `0.9`.' format: float example: 0.95 default: 0.95 minimum: 0.01 maximum: 1 max_tokens: type: integer description: The maximum number of tokens for model output, the GLM-5.1, GLM-5, GLM-4.7, GLM-4.6 series supports 128K maximum output, the GLM-4.5 series supports 96K maximum output, the GLM-4.6v series supports 32K maximum output, the GLM-4.5v series supports 16K maximum output, GLM-4-32B-0414-128K supports 16K maximum output. example: 1024 minimum: 1 maximum: 131072 tool_stream: type: boolean example: false default: false description: Whether to enable streaming response for Function Calls. Default value is false. Only supported by GLM-4.6 and above. Refer the [Stream Tool Call](/guides/tools/stream-tool) tools: type: array description: 'A list of tools the model may call. Currently, only functions are supported as a tool. Use this to provide a list of functions the model may generate JSON inputs for. A max of 128 functions are supported. ' items: anyOf: - $ref: '#/components/schemas/FunctionToolSchema' - $ref: '#/components/schemas/RetrievalToolSchema' - $ref: '#/components/schemas/WebSearchToolSchema' tool_choice: oneOf: - type: string enum: - auto description: Used to control how the model selects which function to call. This is only applicable when the tool type is function. The default value is auto, and only auto is supported. description: Controls how the model selects a tool. stop: type: array description: Stop word list. Generation stops when the model encounters any specified string. Currently, only one stop word is supported, in the format ["stop_word1"]. items: type: string maxItems: 1 response_format: type: object description: 'Specifies the response format of the model. Defaults to text. Supports two formats:{ "type": "text" } plain text mode, returns natural language text, { "type": "json_object" } JSON mode, returns valid JSON data. When using JSON mode, it’s recommended to clearly request JSON output in the prompt.' properties: type: type: string enum: - text - json_object default: text description: 'Output format type: text for plain text, json_object for JSON-formatted output.' required: - type user_id: type: string description: Unique ID for the end user, 6–128 characters. Avoid using sensitive information. minLength: 6 maxLength: 128 ViduFrames2VideoRequest: allOf: - type: object properties: model: type: string description: The model code to be called. enum: - viduq1-start-end - vidu2-start-end prompt: type: string description: Text description of the video, maximum input length of 512 characters. Either image_url or prompt must be provided, or both. image_url: type: array description: 'Images Supports input of two images: the first uploaded image will be treated as the first frame, and the second image as the last frame. The model will use the images provided in this parameter to generate a video. The resolutions of the two input images (first and last frame) must be similar, with the ratio between the resolution of the first frame and the resolution of the last frame falling within `0.8–1.25`. Additionally, the image aspect ratio must be less than `1:4` or `4:1`. Supports image URLs or images encoded in Base64 (ensure accessibility; using image URLs is recommended). Supported formats: `png`, `jpeg`, `.jpg`, `webp`. Image file size must not exceed `50 MB`. Note: After Base64 decoding, the byte length must be less than 50 MB, and the encoding must include the appropriate content type string, such as `data:image/png;base64,{base64_encode}`.' items: type: string minLength: 1 minItems: 1 maxItems: 2 duration: oneOf: - title: viduq1-start-end type: integer description: 'Video duration parameter. Default: `5` , Optional: `5`.' example: 5 enum: - 5 - title: vidu2-start-end type: integer description: 'Video duration parameter. Default: `4` , Optional: `4`.' example: 4 enum: - 4 size: oneOf: - title: viduq1-start-end type: string description: 'Resolution parameter Default: `1920x1080`, Optional: `1920x1080`' example: 1920x1080 enum: - 1920x1080 - title: vidu2-start-end type: string description: 'Resolution parameter Default: `1280x720`, Optional: `1280x720`' example: 1280x720 default: 1280x720 enum: - 1280x720 movement_amplitude: type: string description: 'Motion amplitude Default: `auto` , Optional values: `auto` ,`small` ,`medium` ,`large`' example: auto enum: - auto - small - medium - large with_audio: type: boolean description: Add background music to the generated video. required: - model - $ref: '#/components/schemas/VideoCommonRequest' AsyncImageGenerationResponse: type: object properties: model: type: string description: Model name. task_status: type: string description: 'Processing status, `PROCESSING` (processing), `SUCCESS` (success), `FAIL` (failure). Note: Processing status needs to be obtained via query.' image_result: type: array description: Array containing the generated image URLs. items: type: object properties: url: type: string description: Image URL. The temporary link expires after `30` days, please store it promptly. request_id: type: string description: Unique ID identifying this request, can be submitted by the user or generated by the platform. FunctionObject: type: object properties: name: type: string description: The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64. minLength: 1 maxLength: 64 pattern: ^[a-zA-Z0-9_-]+$ description: type: string description: A description of what the function does, used by the model to choose when and how to call the function. parameters: $ref: '#/components/schemas/FunctionParameters' required: - name - description - parameters VideoCommonRequest: type: object properties: request_id: type: string description: Provided by the client, must be unique; used to distinguish each request’s unique identifier. If not provided by the client, the platform will generate one by default. user_id: type: string description: 'Unique ID of the end-user, assists the platform in intervening in end-user violations, generating illegal or inappropriate information, or other abusive behaviors. ID length requirement: minimum `6` characters, maximum `128` characters.' ChatCompletionResponse: type: object properties: id: description: Task ID type: string request_id: description: Request ID type: string created: description: Request creation time, Unix timestamp in seconds type: integer model: description: Model name type: string choices: type: array description: List of model responses items: type: object properties: index: type: integer description: Result index. message: $ref: '#/components/schemas/ChatCompletionResponseMessage' finish_reason: type: string description: Reason for model inference termination. Can be `stop`, `tool_calls`, `length`, `sensitive`, `model_context_window_exceeded` or `network_error`. usage: type: object description: Token usage statistics returned when the model call ends. properties: prompt_tokens: type: number description: Number of tokens in user input completion_tokens: type: number description: Number of output tokens prompt_tokens_details: type: object properties: cached_tokens: type: number description: Number of tokens served from cache total_tokens: type: integer description: Total number of tokens web_search: description: Search results. type: array items: $ref: '#/components/schemas/WebSearchObjectResponse' ChatCompletionResponseMessageToolCall: type: object properties: function: type: object description: Contains the function name and JSON format parameters generated by the model. properties: name: type: string description: Model-generated function name. arguments: type: object description: JSON format of the function call parameters generated by the model. Validate the parameters before calling the function. required: - name - arguments id: type: string description: Unique identifier for the hit function. type: type: string description: Tool type called by the model, currently only supports ‘function’. WebSearchObjectResponse: type: object properties: title: type: string description: Title. content: type: string description: Content summary. link: type: string description: Result URL. media: type: string description: Website name. icon: type: string description: Website icon. refer: type: string description: Index number. publish_date: type: string description: Website publication date. ChatThinking: type: object description: Only supported by GLM-4.5 series and higher models. This parameter is used to control whether the model enable the chain of thought. properties: type: type: string description: 'Whether to enable the chain of thought(When enabled, GLM-5.1 GLM-5 GLM-5-Turbo GLM-5V-Turbo GLM-4.7 GLM-4.5V will think compulsorily, while GLM-4.6, GLM-4.6V, GLM-4.5 and others will automatically determine whether to think), default: enabled' default: enabled enum: - enabled - disabled clear_thinking: type: boolean description: "Default value is True. Controls whether to clear `reasoning_content` from previous conversation turns. View more in [Thinking Mode](/guides/capabilities/thinking-mode). \n - `true` (default): For this request, the system ignores/removes `reasoning_content` from prior turns, and only keeps non-reasoning context (e.g., user/assistant visible text, tool calls, and tool results). This is recommended for general chat or lightweight tasks to reduce context length and cost. \n - `false`: Retains `reasoning_content` from prior turns and includes it in the context sent to the model. To enable Preserved Thinking, you must forward the full, unmodified, and correctly ordered historical `reasoning_content` in `messages`. Missing, truncated, rewritten, or reordered blocks may degrade performance or prevent the feature from taking effect. \n - Notes: This parameter only affects cross-turn historical thinking blocks; it does not change whether the model generates/returns thinking in the current turn." default: true example: true PageInfo: type: object description: Page dimension information properties: width: type: integer description: Page width example: 600 height: type: integer description: Page height example: 800 required: - width - height parameters: AcceptLanguage: name: Accept-Language in: header schema: type: string description: Config desired response language for HTTP requests. default: en-US,en example: en-US,en enum: - en-US,en required: false securitySchemes: bearerAuth: type: http scheme: bearer description: 'Use the following format for authentication: Bearer [](https://z.ai/manage-apikey/apikey-list)'