> ## Documentation Index > Fetch the complete documentation index at: https://docs.evolink.ai/llms.txt > Use this file to discover all available pages before exploring further. # Doubao Seed 2.0 - Complete API Reference > - Call Doubao Seed 2.0 series models using OpenAI SDK format - Synchronous processing mode, real-time response - **Text Chat**: Single or multi-turn contextual conversation - **System Prompts**: Customize AI role and behavior - **Multimodal Input**: Supports text + image + video mixed input - **Deep Thinking**: Supports enabling chain-of-thought mode for deep reasoning - **Tool Calling**: Function Calling support - **Structured Output**: Supports JSON Object / JSON Schema format output - Quick start? Check the [Quick Start Guide](./doubao-seed-2.0-quickstart) **BaseURL**: The default BaseURL is `https://direct.evolink.ai`, which has better support for text models and long-lived connections. `https://api.evolink.ai` is the primary endpoint for multimodal services and serves as a fallback address for text models. ## OpenAPI ````yaml /en/api-manual/language-series/doubao-seed-2.0/doubao-seed-2.0-reference.json POST /v1/chat/completions openapi: 3.1.0 info: title: Doubao Seed 2.0 Complete API Reference description: >- Complete API reference for Doubao Seed 2.0 chat interface, including all parameters and advanced features license: name: MIT version: 1.0.0 servers: - url: https://direct.evolink.ai description: Production (recommended) - url: https://api.evolink.ai description: Alternative URL security: - bearerAuth: [] tags: - name: Chat Completion description: AI chat completion related APIs paths: /v1/chat/completions: post: tags: - Chat Completion summary: Doubao Seed 2.0 Chat API description: >- - Call Doubao Seed 2.0 series models using OpenAI SDK format - Synchronous processing mode, real-time response - **Text Chat**: Single or multi-turn contextual conversation - **System Prompts**: Customize AI role and behavior - **Multimodal Input**: Supports text + image + video mixed input - **Deep Thinking**: Supports enabling chain-of-thought mode for deep reasoning - **Tool Calling**: Function Calling support - **Structured Output**: Supports JSON Object / JSON Schema format output - Quick start? Check the [Quick Start Guide](./doubao-seed-2.0-quickstart) operationId: createChatCompletionDoubaoSeed20 requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/ChatCompletionRequest' examples: simple_text: summary: Simple text chat value: model: doubao-seed-2.0-pro messages: - role: user content: Please introduce yourself multi_turn: summary: Multi-turn conversation (context understanding) value: model: doubao-seed-2.0-pro messages: - role: user content: What is Python? - role: assistant content: Python is a high-level programming language... - role: user content: What are its advantages? system_prompt: summary: Using system prompt value: model: doubao-seed-2.0-pro messages: - role: system content: >- You are a professional Python programming assistant, answer questions concisely. - role: user content: How to read a file? vision: summary: Multimodal input (text + image) value: model: doubao-seed-2.0-pro messages: - role: user content: - type: text text: >- Please describe in detail the scene and main elements in this image. - type: image_url image_url: url: https://example.com/image.png thinking_mode: summary: Enable deep thinking mode value: model: doubao-seed-2.0-pro messages: - role: user content: Prove that sqrt(2) is irrational thinking: type: enabled max_completion_tokens: 16384 code_generation: summary: Using code-specialized model value: model: doubao-seed-2.0-code messages: - role: system content: You are an advanced programming assistant. - role: user content: Implement a quicksort algorithm in Python responses: '200': description: Chat completion successful content: application/json: schema: $ref: '#/components/schemas/ChatCompletionResponse' '400': description: Invalid request parameters content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 400 message: Invalid request parameters type: invalid_request_error '401': description: Unauthenticated, invalid or expired token content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 401 message: Invalid or expired token type: authentication_error '402': description: Insufficient quota, top-up required content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 402 message: Insufficient quota type: insufficient_quota_error fallback_suggestion: https://evolink.ai/dashboard/billing '403': description: Access denied content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 403 message: Access denied for this model type: permission_error param: model '404': description: Resource not found content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 404 message: Specified model not found type: not_found_error param: model fallback_suggestion: doubao-seed-2.0-pro '413': description: Request body too large content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 413 message: Image file too large type: request_too_large_error param: content fallback_suggestion: compress image to under 10MB '429': description: Rate limit exceeded content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 429 message: Rate limit exceeded type: rate_limit_error fallback_suggestion: retry after 60 seconds '500': description: Internal server error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 500 message: Internal server error type: internal_server_error fallback_suggestion: try again later '502': description: Upstream service error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 502 message: Upstream AI service unavailable type: upstream_error fallback_suggestion: try different model '503': description: Service temporarily unavailable content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: code: 503 message: Service temporarily unavailable type: service_unavailable_error fallback_suggestion: retry after 30 seconds components: schemas: ChatCompletionRequest: type: object required: - model - messages properties: model: type: string description: >- Chat model name - `doubao-seed-2.0-pro`: Flagship, strongest overall capability, ideal for complex reasoning and high-quality generation - `doubao-seed-2.0-lite`: Lightweight, faster speed, cost-effective - `doubao-seed-2.0-mini`: Ultra-fast, quickest response, suitable for simple tasks - `doubao-seed-2.0-code`: Code-specialized, optimized for code generation and understanding enum: - doubao-seed-2.0-pro - doubao-seed-2.0-lite - doubao-seed-2.0-mini - doubao-seed-2.0-code default: doubao-seed-2.0-pro example: doubao-seed-2.0-pro messages: type: array description: >- List of conversation messages, supports multi-turn conversation and multimodal input (text, image, video) items: $ref: '#/components/schemas/Message' minItems: 1 thinking: type: object description: |- Control whether the model enables deep thinking mode Different models may vary in support and default values properties: type: type: string description: >- Thinking mode - `enabled`: Enable thinking mode, model must think before answering - `disabled`: Disable thinking mode, model answers directly without thinking - `auto`: Auto thinking mode, model decides whether to think based on the question enum: - enabled - disabled - auto example: enabled required: - type stream: type: boolean description: >- Whether to stream the response content - `false`: Model generates all content before returning the result at once - `true`: Returns model-generated content incrementally via SSE protocol, ending with a `data: [DONE]` message. When stream is true, you can set the stream_options field to get token usage statistics default: false example: false stream_options: type: object nullable: true description: Options for streaming responses. Can be set when stream is true properties: include_usage: type: boolean nullable: true description: >- Whether to output token usage information before streaming output ends - `true`: Before the `data: [DONE]` message, an additional chunk will be returned with the usage field containing total token usage for the entire request, and an empty choices array - `false`: No chunk is returned with token usage information before output ends default: false example: true chunk_include_usage: type: boolean nullable: true description: >- Whether each chunk in streaming output includes cumulative token usage information up to that point - `true`: The usage field in each returned chunk contains cumulative token usage from the start of the request to the current chunk - `false`: Token usage information is not included in each chunk default: false max_tokens: type: integer nullable: true description: >- Maximum length of model response (in tokens) **Note**: - Model response does not include chain-of-thought content (model response = model output - model chain-of-thought) - Total output token length is also limited by the model's context length - Cannot be set simultaneously with max_completion_tokens default: 4096 example: 4096 max_completion_tokens: type: integer nullable: true description: >- Controls the maximum output length of the model, including both response and chain-of-thought content (in tokens) **Note**: - Value range: [0, 65536] - When configured, the default value of max_tokens becomes ineffective; the model outputs content (response and chain-of-thought) as needed until reaching this value - Cannot be set simultaneously with max_tokens - Recommended when deep thinking mode is enabled minimum: 0 maximum: 65536 example: 16384 temperature: type: number nullable: true description: >- Sampling temperature, controls output randomness **Note**: - Value range: [0, 2] - Lower values (e.g., 0.2): More deterministic, more focused output - Higher values (e.g., 0.8): More random, more creative output - A value of 0 means the model only considers the token with the highest log probability - It is recommended to adjust only temperature or top_p, not both minimum: 0 maximum: 2 default: 1 example: 0.7 top_p: type: number nullable: true description: >- Nucleus sampling probability threshold **Note**: - Value range: [0, 1] - The model considers tokens within the top_p probability mass - 0.1 means only considering the top 10% of tokens by probability mass - Higher values produce more random output, lower values produce more deterministic output - It is recommended to adjust only temperature or top_p, not both minimum: 0 maximum: 1 default: 0.7 example: 0.9 stop: description: >- The model will stop generating when it encounters the string(s) specified in the stop field. The stop string itself will not be included in the output. Up to 4 strings are supported **Note**: Deep thinking models do not support this field nullable: true oneOf: - type: string - type: array items: type: string maxItems: 4 example: - hello - weather reasoning_effort: type: string nullable: true description: >- Limits the amount of thinking effort, reducing thinking depth can improve speed and consume fewer tokens - `minimal`: Disable thinking, answer directly - `low`: Lightweight thinking, prioritizes quick response - `medium`: Balanced mode, balances speed and depth - `high`: Deep analysis, handles complex problems enum: - minimal - low - medium - high default: medium example: medium response_format: type: object description: |- Specify the model response format Supports three formats: text (default), json_object, json_schema properties: type: type: string description: >- Response format type - `text`: Default text format - `json_object`: Model response content is organized as a JSON object structure - `json_schema`: Model response content follows the JSON structure defined in the schema field enum: - text - json_object - json_schema default: text example: text json_schema: type: object description: JSON structure definition (required when type is json_schema) properties: name: type: string description: User-defined JSON structure name description: type: string nullable: true description: >- Description of the response purpose; the model will use this to determine how to respond in the specified format schema: type: object description: >- JSON format definition for the response, described as a JSON Schema object strict: type: boolean nullable: true description: >- Whether to enable strict adherence mode - `true`: The model will always strictly follow the format defined in the schema field - `false`: The model will try its best to follow the structure defined in the schema field default: false required: - name - schema required: - type frequency_penalty: type: number nullable: true description: >- Frequency penalty coefficient **Note**: - Value range: [-2.0, 2.0] - Positive values penalize new tokens based on their frequency in the text so far, reducing the likelihood of the model repeating the same content verbatim minimum: -2 maximum: 2 default: 0 example: 0 presence_penalty: type: number nullable: true description: >- Presence penalty coefficient **Note**: - Value range: [-2.0, 2.0] - Positive values penalize new tokens based on whether they have appeared in the text so far, increasing the likelihood of the model talking about new topics minimum: -2 maximum: 2 default: 0 example: 0 logprobs: type: boolean nullable: true description: >- Whether to return log probabilities of output tokens - `false`: Do not return log probability information - `true`: Return log probabilities for each output token in the message content **Note**: Deep thinking models do not support this field default: false top_logprobs: type: integer nullable: true description: >- Specify the number of most likely tokens to return at each output token position, each with an associated log probability - Value range: [0, 20] - Can only be set when logprobs is true **Note**: Deep thinking models do not support this field minimum: 0 maximum: 20 default: 0 logit_bias: type: object nullable: true description: >- Adjust the probability of specified tokens appearing in the model output **Note**: - Accepts a map where keys are token IDs from the vocabulary and values are bias values - Bias value range: [-100, 100] - -1 reduces the likelihood of selection, 1 increases the likelihood of selection - -100 completely prohibits selection of the token, 100 causes only that token to be selectable **Note**: Deep thinking models do not support this field additionalProperties: type: integer minimum: -100 maximum: 100 tools: type: array nullable: true description: >- List of tools to be called; the model response may contain tool call requests items: $ref: '#/components/schemas/Tool' parallel_tool_calls: type: boolean description: >- Whether the model response is allowed to contain multiple tool calls for this request - `true`: Allow returning multiple tool calls - `false`: The number of tool calls returned is <= 1 default: true tool_choice: description: >- Whether the model response should contain tool calls for this request **String mode**: - `none`: Model response does not contain tool calls - `required`: Model response must contain tool calls - `auto`: Model decides whether to include tool calls (default when tools are provided) **Object mode**: Specify the scope of tools to be called oneOf: - type: string enum: - none - auto - required - $ref: '#/components/schemas/ToolChoiceObject' ChatCompletionResponse: type: object properties: id: type: string description: Unique identifier for this request example: 0217714854126607f5a9cf8ed5b018c76e4ad3dc2810db57ffb50 model: type: string description: Actual model name and version used for this request example: doubao-seed-2-0-pro-260215 object: type: string enum: - chat.completion description: Response type, always chat.completion example: chat.completion service_tier: type: string description: |- Service tier for this request - `default`: Default service tier - `scale`: Used reserved capacity quota enum: - default - scale example: default created: type: integer description: Unix timestamp (in seconds) of when this request was created example: 1771485416 choices: type: array description: Model output content for this request items: $ref: '#/components/schemas/Choice' usage: $ref: '#/components/schemas/Usage' ErrorResponse: type: object properties: error: type: object properties: code: type: integer description: HTTP status error code message: type: string description: Error description message type: type: string description: Error type param: type: string description: Related parameter name fallback_suggestion: type: string description: Suggestion on error Message: type: object required: - role - content properties: role: type: string description: |- Message role - `user`: User message - `assistant`: AI assistant message (for multi-turn conversations) - `system`: System prompt (define AI role and behavior) - `tool`: Tool response message (for tool calling scenarios) enum: - user - assistant - system - tool example: user content: type: array description: >- Message content. Supports two formats: **1. Plain text string**: You can pass a string directly, e.g., `"content": "Please introduce yourself"` **2. Object array** (supports multimodal input including text, image, video): See the structure below items: $ref: '#/components/schemas/ContentPart' reasoning_content: type: string description: >- Chain-of-thought content in model messages (assistant role only) Only used when passing historical chain-of-thought content in multi-turn conversations tool_calls: type: array description: Tool call part of model messages (assistant role only) items: $ref: '#/components/schemas/ToolCall' tool_call_id: type: string description: >- Tool call request ID generated by the model (required for tool role only) Used to associate tool return results with model requests, avoiding confusion in multi-tool call scenarios Tool: type: object required: - type - function properties: type: type: string description: Tool type, currently only supports function enum: - function function: type: object required: - name description: Function definition properties: name: type: string description: Name of the function to call description: type: string description: >- Description of the function; the model will use this to determine whether to call this tool parameters: type: object description: >- Function request parameters, described in JSON Schema format **Note**: - All field names are case-sensitive - parameters must be a valid JSON Schema object - It is recommended to use English field names and put Chinese descriptions in the description field ToolChoiceObject: type: object required: - type - function description: Specify the scope of tools to be called properties: type: type: string description: Call type enum: - function function: type: object required: - name properties: name: type: string description: Name of the tool to be called Choice: type: object properties: index: type: integer description: Index of the current element in the choices list example: 0 message: $ref: '#/components/schemas/AssistantMessage' finish_reason: type: string description: >- Reason why the model stopped generating tokens - `stop`: Model output ended naturally, or was truncated due to hitting a string specified in the stop parameter - `length`: Model output was truncated due to reaching max_tokens / max_completion_tokens / context length limit - `content_filter`: Model output was blocked by content moderation - `tool_calls`: Model called a tool enum: - stop - length - content_filter - tool_calls example: stop logprobs: type: object nullable: true description: Log probability information for the current content properties: content: type: array nullable: true description: Token log probability information for each content element items: $ref: '#/components/schemas/LogprobContent' moderation_hit_type: type: string nullable: true description: >- When model output contains sensitive content, returns the matched risk classification label - `severe_violation`: Model output involves severe violation - `violence`: Model output involves violent behavior enum: - severe_violation - violence Usage: type: object description: Token usage for this request properties: total_tokens: type: integer description: Total number of tokens consumed by this request (input + output) example: 271 prompt_tokens: type: integer description: Number of tokens in the content input to the model example: 15 prompt_tokens_details: type: object description: Details of input token count properties: cached_tokens: type: integer description: Token usage for cached input content example: 0 completion_tokens: type: integer description: Tokens consumed by model output content example: 256 completion_tokens_details: type: object description: Details of output token consumption properties: reasoning_tokens: type: integer description: Number of tokens consumed by chain-of-thought output example: 0 ContentPart: oneOf: - $ref: '#/components/schemas/TextContent' - $ref: '#/components/schemas/ImageContent' - $ref: '#/components/schemas/VideoContent' ToolCall: type: object properties: id: type: string description: ID of the called tool, generated by the model type: type: string description: Tool type, currently only supports function enum: - function function: type: object properties: name: type: string description: Name of the function to call arguments: type: string description: >- Arguments for the function to call, in JSON format **Note**: The model does not always generate valid JSON and may hallucinate parameters that are not defined. It is recommended to validate parameters before calling the function AssistantMessage: type: object properties: role: type: string description: Role of the content output, always assistant enum: - assistant example: assistant content: type: string description: Message content generated by the model example: >- Hello! Doubao Seed 2.0 is a next-generation large language model from ByteDance, featuring enhanced reasoning, multimodal understanding, and deep thinking capabilities. reasoning_content: type: string nullable: true description: |- Chain-of-thought content from the model's problem-solving process Only returned when deep thinking mode is enabled tool_calls: type: array nullable: true description: Tool calls generated by the model items: $ref: '#/components/schemas/ToolCall' LogprobContent: type: object properties: token: type: string description: Current token bytes: type: array nullable: true description: UTF-8 value of the current token, as a list of integers items: type: integer logprob: type: number description: Log probability of the current token top_logprobs: type: array description: >- List of the most likely tokens at the current token position and their log probabilities items: type: object properties: token: type: string description: Current token bytes: type: array nullable: true description: UTF-8 value of the current token items: type: integer logprob: type: number description: Log probability of the current token TextContent: title: Text Content type: object required: - type - text properties: type: type: string enum: - text description: Content type text: type: string description: Text content example: Please describe this image in detail ImageContent: title: Image Content type: object required: - type - image_url properties: type: type: string enum: - image_url description: Content type image_url: type: object required: - url properties: url: type: string format: uri description: |- Image URL or Base64 encoding **Supported formats**: - Image link (publicly accessible) - Base64 encoded image example: https://example.com/image.png detail: type: string nullable: true description: |- Image understanding detail level - `low`: Low detail - `high`: High detail - `xhigh`: Ultra-high detail enum: - low - high - xhigh VideoContent: title: Video Content type: object required: - type - video_url properties: type: type: string enum: - video_url description: Content type video_url: type: object required: - url properties: url: type: string format: uri description: >- Video URL or Base64 encoding **Supported formats**: - Video link (publicly accessible) - Base64 encoded video **Note**: Audio content within the video is not supported for understanding example: https://example.com/video.mp4 fps: type: number nullable: true description: >- Frame extraction rate - Value range: [0.2, 5] - Higher values make the model more sensitive to visual changes in the video - Lower values make the model less sensitive to visual changes, but consume fewer tokens and are faster minimum: 0.2 maximum: 5 default: 1 example: 1 securitySchemes: bearerAuth: type: http scheme: bearer description: >- ##All APIs require Bearer Token authentication## **Get API Key:** Visit the [API Key Management Page](https://evolink.ai/dashboard/keys) to get your API Key **Add to request headers:** ``` Authorization: Bearer YOUR_API_KEY ``` ````