> ## Documentation Index > Fetch the complete documentation index at: https://docs.totoy.ai/llms.txt > Use this file to discover all available pages before exploring further. # Creates a Knowledge Base Chat Response for the given Messages. ## OpenAPI ````yaml /api-reference/openapi.yaml post /knowledge-bases/{knowledge_base_id}/chat openapi: 3.0.3 info: title: Totoy API description: >- Explain documents in simple, plain or detailed language and create Knowledge Bases from your documents in 19 languages. termsOfService: https://www.totoy.ai contact: name: Totoy Support email: support@totoy.ai license: name: MIT url: https://raw.githubusercontent.com/totoy-ai/totoy-openapi/main/LICENSE version: 1.1.5 servers: - url: https://api.totoy.ai/v1 security: - ApiKeyAuth: [] tags: - name: Explanation description: >- Given a document in any language, an output language and one of three language levels, the assistant will return an `Explanation` for this document. - name: Knowledge Bases description: >- Create `Knowledge Bases` from document or text `Sources` and chat with an assistant about them in 19 languages. - name: Sources description: >- `Sources` are used to upload documents (.pdf, .jpg or .png) or plain texts that can be used with features like `Explanation` or `Knowledge Bases.` - name: Projects description: '`Projects` are used to group `Knowledge Bases` and `Explanations`.' - name: Organization description: Get information about your `Organization`. paths: /knowledge-bases/{knowledge_base_id}/chat: post: tags: - Knowledge Bases summary: Creates a Knowledge Base Chat Response for the given Messages. operationId: chatWithKnowledgeBase parameters: - $ref: '#/components/parameters/knowledgeBaseIdParam' requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/KnowledgeBaseChatRequest' examples: knowledge-base-chat-request-example-1: $ref: '#/components/examples/knowledge-base-chat-request-example-1' knowledge-base-chat-request-example-2: $ref: '#/components/examples/knowledge-base-chat-request-example-2' responses: '200': description: OK content: application/json: schema: $ref: '#/components/schemas/KnowledgeBaseChatResponse' examples: knowledge-base-chat-response-example-1: $ref: '#/components/examples/knowledge-base-chat-response-example-1' knowledge-base-chat-response-example-2: $ref: '#/components/examples/knowledge-base-chat-response-example-2' 4XX: $ref: '#/components/responses/general4XXResponse' 5XX: $ref: '#/components/responses/general5XXResponse' components: parameters: knowledgeBaseIdParam: name: knowledge_base_id in: path required: true schema: type: string maxLength: 28 pattern: ^kb_[a-zA-Z0-9]{25}$ description: Unique identifier for a `Knowledge Base`. examples: knowledge-base-id-example: $ref: '#/components/examples/knowledge-base-id-example' schemas: KnowledgeBaseChatRequest: title: KnowledgeBaseChatRequest type: object additionalProperties: false properties: output_language: $ref: '#/components/schemas/OutputLanguage' language_level: $ref: '#/components/schemas/LanguageLevel' messages: description: >- The conversation history. Provide at least one user message to start the conversation. type: array items: $ref: '#/components/schemas/Message' maxItems: 50 markdown_response: description: >- Whether the response should be returned as Markdown formatted text. If 'false', the response will be returned in plain text. type: boolean default: true example: true required: - output_language - language_level - messages KnowledgeBaseChatResponse: title: KnowledgeBaseChatResponse type: object additionalProperties: false properties: project_id: description: The `Project` that the `Knowledge Base` is assigned to. maxLength: 28 pattern: ^pj_[a-zA-Z0-9]{25}$ type: string example: pj_avX7imfLaPcQnv5ckvGlOEBA9 output_language: $ref: '#/components/schemas/OutputLanguage' language_level: $ref: '#/components/schemas/LanguageLevel' answer: $ref: '#/components/schemas/Message' markdown_response: description: Whether the response is a Markdown formatted text. type: boolean example: true knowledge_base_id: description: >- The unique identifier of the `Knowledge Base` that was used for the `Chat`. maxLength: 28 pattern: ^kb_[a-zA-Z0-9]{25}$ type: string example: kb_feMfJbeqbAEj4u8K5HqmKpUbY required: - project_id - output_language - language_level - answer - markdown_response - knowledge_base_id OutputLanguage: title: OutputLanguage description: > The ISO 639-1 code for the language in which the `Explanation` should be generated. Supported Languages: - ar: Arabic - bs: Bosnian - cs: Czech - de: German - es: Spanish - en: English - fa: Farsi - fr: French - hr: Croatian - hu: Hungarian - it: Italian - pl: Polish - ro: Romanian - sk: Slovak - sl: Slovenian - sr: Serbian - tl: Tagalog - tr: Turkish - uk: Ukrainian - zh: Mandarin type: string enum: - ar - bs - cs - de - es - en - fa - fr - hr - hu - it - pl - ro - sk - sl - sr - tl - tr - uk - zh example: en LanguageLevel: title: LanguageLevel description: > One of three language levels for the generated text: 1. Simple Language: Answers are short and simple, using easy words and a clear structure. 2. Plain Language: Answers are short and the assistant explains complicated terms. 3. Detailed Language: Answers are not simplified and the assistant answers in detail. type: string enum: - simple - plain - detailed example: plain Message: title: Message description: > Represents a message within an `Explanation` or a `Knowledge Base Chat`. For `Explanations` - An `assistant` message is either an initial `Explanation` for the document when no messages were provided in the request or an answer to a user message in the provided conversation history. type: object additionalProperties: false properties: role: description: The entity that produced the message. One of `user` or `assistant`. type: string enum: - user - assistant content: description: The content of the message. The maximum length is 16,000 characters. type: string maxLength: 16000 references: type: array items: $ref: '#/components/schemas/Reference' maxItems: 50 required: - role - content example: role: assistant content: >- Hi! The co-founders of Totoy are Benedikt Hielscher, Francis Rafal, Marcel Koller, Michael Perger and Simon Hoffmann.[R1] Problem: title: Problem description: A Response Body for HTTP Problem Details in RFC 9457 format. type: object additionalProperties: false properties: type: description: A URI reference that identifies the problem type. type: string maxLength: 512 example: https://docs.totoy.ai/probs/out-of-credit title: description: A short, human-readable summary of the problem type. type: string maxLength: 512 example: You do not have enough credit. status: description: >- The HTTP status code generated by the origin server for this occurrence of the problem. type: integer format: int32 example: 403 detail: description: >- A human-readable explanation specific to this occurrence of the problem. type: string maxLength: 4096 example: You do not have enough credit to perform this operation. instance: description: >- A URI reference that identifies the specific occurrence of the problem. type: string maxLength: 512 example: /projects/pj_avX7imfLaPcQnv5ckvGlOEBA9 Reference: title: Reference description: > Represents a reference to a `Source` that was used to generate the answer. The reference is used to provide backlinks to the original source and to provide context for the generated answer. type: object additionalProperties: false properties: text: description: String in the message content that needs to be replaced. type: string maxLength: 8 example: '[R1]' source_id: description: What `Source` the generated answer depended on. maxLength: 29 pattern: ^src_[a-zA-Z0-9]{25}$ type: string page_number: description: >- The page number in the `Source` the generated answer depended on. Only applicable for documents. type: integer format: int32 example: 5 nullable: true minimum: 1 maximum: 1000 backlink: description: The backlink of the `Source`, so it can be called from the client. type: string maxLength: 512 example: https://www.example.com/text.pdf custom_metadata: $ref: '#/components/schemas/CustomMetadata' required: - text - source_id example: text: '[R1]' source_id: src_oXiYDg8QOc9JsOk86idRXiywD page_number: 5 backlink: https://www.totoy.ai/totoy-company-overview.pdf custom_metadata: internal_tag: general_documents CustomMetadata: title: CustomMetadata description: >- Custom optional metadata for a `Source` provided by a client. Up to 10 key-value pairs. type: object additionalProperties: true maxProperties: 10 example: internal_tag: general_documents examples: knowledge-base-chat-request-example-1: summary: Knowledge Base Chat Request Example 1 description: >- An example for requesting a plain English answer to a question to a `Knowledge Base`. value: output_language: en language_level: plain markdown_response: true messages: - role: user content: Who are the co-founders of Totoy? knowledge-base-chat-request-example-2: summary: Knowledge Base Chat Request Example 2 description: >- An example for requesting a plain English answer to a follow-up question to a `Knowledge Base`, based on the response to a previous question. value: output_language: en language_level: plain messages: - role: user content: Who are the co-founders of Totoy? - role: assistant content: >- Hi! The co-founders of Totoy are Benedikt Hielscher, Francis Rafal, Marcel Koller, Michael Perger and Simon Hoffmann.[R1] references: - text: '[R1]' source_id: src_oXiYDg8QOc9JsOk86idRXiywD page_number: 5 backlink: https://www.totoy.ai/totoy-company-overview.pdf custom_metadata: internal_tag: general_documents - role: user content: What are they trying to achieve? markdown_response: true knowledge-base-chat-response-example-1: summary: Knowledge Base Chat Response Example 1 description: >- An example for a generated plain English answer to a question to a `Knowledge Base`. value: project_id: pj_avX7imfLaPcQnv5ckvGlOEBA9 output_language: en language_level: plain answer: role: assistant content: >- Hi! The co-founders of Totoy are Benedikt Hielscher, Francis Rafal, Marcel Koller, Michael Perger and Simon Hoffmann.[R1] references: - text: '[R1]' source_id: src_oXiYDg8QOc9JsOk86idRXiywD page_number: 5 backlink: https://www.totoy.ai/totoy-company-overview.pdf custom_metadata: internal_tag: general_documents knowledge_base_id: kb_feMfJbeqbAEj4u8K5HqmKpUbY markdown_response: true knowledge-base-chat-response-example-2: summary: Knowledge Base Chat Response Example 2 description: >- An example for a generated plain English answer to a follow-up question to a `Knowledge Base` from Knowledge Base Chat Request Example 2. value: project_id: pj_avX7imfLaPcQnv5ckvGlOEBA9 output_language: en language_level: plain answer: role: assistant content: >- They want to solve functional illiteracy[R1]. They want to achieve that by making complicated documents understandable for everyone.[R2] references: - text: '[R1]' source_id: src_oXiYDg8QOc9JsOk86idRXiywD page_number: 1 backlink: https://www.totoy.ai/totoy-company-overview.pdf custom_metadata: internal_tag: general_documents - text: '[R2]' source_id: src_oXiYDg8QOc9JsOk86idRXiywD page_number: 2 backlink: https://www.totoy.ai/totoy-company-overview.pdf custom_metadata: internal_tag: general_documents knowledge_base_id: kb_feMfJbeqbAEj4u8K5HqmKpUbY markdown_response: true knowledge-base-id-example: summary: Knowledge Base ID Example description: An example for a `knowledge_base_id`. value: kb_feMfJbeqbAEj4u8K5HqmKpUbY responses: general4XXResponse: description: Client Error content: application/problem+json: schema: $ref: '#/components/schemas/Problem' general5XXResponse: description: Server Error content: application/problem+json: schema: $ref: '#/components/schemas/Problem' securitySchemes: ApiKeyAuth: type: http scheme: bearer ````