API Reference

Chat

Chat with Model (Legacy)

To chat with a large language model, without using your data as context, call POST /v1.0/tenants/[tenant-guid]/assistant/rag to the Assistant service, which by default listens on port 8000. Submit a request body containing the following parameters:

  • Question string the input query to be processed by the assistant
  • EmbeddingModel string the model used for generating embeddings (e.g. sentence-transformers/all-MiniLM-L6-v2)
  • MaxResults number the maximum number of results to retrieve from the vector database
  • VectorDatabaseName string name of the vector database
  • VectorDatabaseTable string table or view name used for querying vector data
  • VectorDatabaseHostname string hostname of the vector database server
  • VectorDatabasePort number port number of the vector database server
  • VectorDatabaseUser string username used to authenticate with the vector database
  • VectorDatabasePassword string password used for database authentication
  • GenerationProvider string provider used for text generation (e.g. ollama)
  • GenerationApiKey string API key for the text generation provider
  • GenerationModel string the model used for generating responses (e.g. qwen2.5:7b)
  • HuggingFaceApiKey string API key for Hugging Face models (if applicable)
  • Temperature number controls randomness in the generated text (higher values = more creativity)
  • MaxTokens number maximum number of tokens allowed in the generated response
  • Stream boolean whether to stream generated tokens in real time
  • OllamaHostname string hostname of the Ollama generation service
  • OllamaPort number port number for the Ollama generation service
  • TopP number nucleus sampling parameter for token generation
  • PromptPrefix string additional prompt text to influence tone or style of the response
  • ContextSort boolean whether to sort retrieved context entries
  • SortByMaxSimilarity boolean whether to sort context entries based on maximum similarity score
  • ContextScope number number of top context entries to include (0 = none)
  • Rerank boolean whether to apply reranking to retrieved results
  • RerankModel string model used for reranking (e.g. cross-encoder/ms-marco-MiniLM-L-6-v2)
  • RerankTopK number number of top results to rerank
curl --location 'http://view.homedns.org:8000/v1.0/tenants/00000000-0000-0000-0000-000000000000/assistant/rag' \
--header 'Cache-Control: no-cache' \
--header 'Content-Type: application/json' \
--data '{
    "Question": "What information do you have?",
    "EmbeddingModel": "sentence-transformers/all-MiniLM-L6-v2",
    "MaxResults": 10,
    "VectorDatabaseName": "vectordb",
    "VectorDatabaseTable": "minilm",
    "VectorDatabaseHostname": "pgvector",
    "VectorDatabasePort": 5432,
    "VectorDatabaseUser": "postgres",
    "VectorDatabasePassword": "password",
    "GenerationProvider": "ollama",
    "GenerationApiKey": "",
    "GenerationModel": "qwen2.5:7b",
    "HuggingFaceApiKey": "",
    "Temperature": 0.1,
    "MaxTokens": 75,
    "Stream": false,
    "OllamaHostname": "ollama",
    "OllamaPort": 11434,
    "TopP": 0.95,
    "PromptPrefix": "talk like a pirate",
    "ContextSort": true,
    "SortByMaxSimilarity": true,
    "ContextScope": 0,
    "Rerank": false,
    "RerankModel": "cross-encoder/ms-marco-MiniLM-L-6-v2",
    "RerankTopK": 5
}'

import { ViewAssistantSdk } from "view-sdk";

const assistant = new ViewAssistantSdk(
  "00000000-0000-0000-0000-000000000000", //tenant Id
  "default", //access token
  "http://localhost:8000/" //endpoint
);

const chatRagQuestion_LEGACY = async () => {
  try {
    const response = await assistant.chatRagQuestion_LEGACY(
      {
        Question: "What information do you have?",
        EmbeddingModel: "sentence-transformers/all-MiniLM-L6-v2",
        MaxResults: 10,
        VectorDatabaseName: "vectordb",
        VectorDatabaseTable: "minilm",
        VectorDatabaseHostname: "pgvector",
        VectorDatabasePort: 5432,
        VectorDatabaseUser: "postgres",
        VectorDatabasePassword: "password",
        GenerationProvider: "ollama",
        GenerationApiKey: "",
        GenerationModel: "qwen2.5:7b",
        HuggingFaceApiKey: "",
        Temperature: 0.1,
        MaxTokens: 75,
        Stream: false,
        OllamaHostname: "ollama",
        OllamaPort: 11434,
        TopP: 0.95,
        PromptPrefix: "talk like a pirate",
        ContextSort: true,
        SortByMaxSimilarity: true,
        ContextScope: 0,
        Rerank: false,
        RerankModel: "cross-encoder/ms-marco-MiniLM-L-6-v2",
        RerankTopK: 5,
      },
      (token) => {
        console.log(token);
      }
    );
    console.log(response);
  } catch (err) {
    console.log("Error in chatRagQuestion_LEGACY:", err);
  }
};

chatRagQuestion_LEGACY();

The response will be sent using chunked transfer-encoding and a content-type of text/event-stream, meaning each chunk in the response will be encoded as an event. An example response, using curl -v --raw is as follows:

Your HTTP client should use chunked transfer encoding and deserialize each line beginning with data: as a payload line. If the string that follows data: is deserializable to JSON, the token property can be extracted and appended to the resultant display. Refer to the View C# SDK for Assistant for more details.

RAG Messages

The retrieval augmented generation (RAG) API follows the same syntax as the chat API, but uses a separate endpoint and a request body with more properties. The endpoint for the RAG API is POST /v1.0/tenants/[tenant-guid]/assistant/rag/chat and the request body has the following structure:

curl --location 'http://view.homedns.org:8000/v1.0/tenants/00000000-0000-0000-0000-000000000000/assistant/rag/chat' \
--header 'Cache-Control: no-cache' \
--header 'Content-Type: application/json' \
--data '{
    "Messages": [ {"role": "user", "content": "Do you have Q3 luxetech financials?"},
                {"role": "assistant", "content": "Unfortunately I do not have context on any documents related to Q3 luxetech financials."},
                {"role": "user", "content": "Are you sure you dont have them?"}
                ],
    "EmbeddingModel": "sentence-transformers/all-MiniLM-L6-v2",
    "MaxResults": 10,
    "VectorDatabaseName": "vectordb",
    "VectorDatabaseTable": "minilm",
    "VectorDatabaseHostname": "pgvector",
    "VectorDatabasePort": 5432,
    "VectorDatabaseUser": "postgres",
    "VectorDatabasePassword": "password",
    "GenerationProvider": "ollama",
    "GenerationApiKey": "",
    "GenerationModel": "qwen2.5:7b",
    "HuggingFaceApiKey": "",
    "Temperature": 0.1,
    "TopP": 0.95,
    "MaxTokens": 75,
    "Stream": true,
    "OllamaHostname": "ollama",
    "OllamaPort": 11434,
    "PromptPrefix": "",
    "ContextSort": true,
    "SortByMaxSimilarity": true,
    "ContextScope": 0,
    "Rerank": false,
    "RerankModel": "cross-encoder/ms-marco-MiniLM-L-6-v2",
    "RerankTopK": 5
}'

import { ViewAssistantSdk } from "view-sdk";

const assistant = new ViewAssistantSdk(
  "00000000-0000-0000-0000-000000000000", //tenant Id
  "default", //access token
  "http://localhost:8000/" //endpoint
);

const chatRagMessages = async () => {
  try {
    const response = await assistant.chatRagMessages(
      {
        Messages: [
          { role: "user", content: "Do you have Q3 luxetech financials?" },
          {
            role: "assistant",
            content:
              "Unfortunately I do not have context on any documents related to Q3 luxetech financials.",
          },
          { role: "user", content: "Are you sure you dont have them?" },
        ],
        EmbeddingModel: "sentence-transformers/all-MiniLM-L6-v2",
        MaxResults: 10,
        VectorDatabaseName: "vectordb",
        VectorDatabaseTable: "minilm",
        VectorDatabaseHostname: "pgvector",
        VectorDatabasePort: 5432,
        VectorDatabaseUser: "postgres",
        VectorDatabasePassword: "password",
        GenerationProvider: "ollama",
        GenerationApiKey: "",
        GenerationModel: "qwen2.5:7b",
        HuggingFaceApiKey: "",
        Temperature: 0.1,
        TopP: 0.95,
        MaxTokens: 75,
        Stream: true,
        OllamaHostname: "ollama",
        OllamaPort: 11434,
        PromptPrefix: "",
        ContextSort: true,
        SortByMaxSimilarity: true,
        ContextScope: 0,
        Rerank: false,
        RerankModel: "cross-encoder/ms-marco-MiniLM-L-6-v2",
        RerankTopK: 5,
      },
      (token) => {
        console.log(token);
      }
    );
    console.log(response);
  } catch (err) {
    console.log("Error in chatRagQuestion_LEGACY:", err);
  }
};

chatRagMessages();

Similar to the chat API, the RAG API will return a result using chunked transfer encoding and a content-type of text/event-stream, meaning your HTTP client should account for these.

Assistant Config Chat

To chat a particular chat config, call POST: v1.0/tenants/{{tenant-guid}}/assistant/chat/[config-guid]

curl --location 'http://view.homedns.org:8000/v1.0/tenants/00000000-0000-0000-0000-000000000000/assistant/chat/578b0872-8186-46b7-bfa3-1871155f4e3a' \
--header 'Cache-Control: no-cache' \
--header 'Content-Type: application/json' \
--data '{
    "messages": [ {"role": "user", "content": "Do you have Q3 luxetech financials?"},
                {"role": "assistant", "content": "Unfortunately I do not have context on any documents related to Q3 luxetech financials."},
                {"role": "user", "content": "Are you sure you dont have them?"}
                ],
    "stream": false
}
'

import { ViewAssistantSdk } from "view-sdk";

const assistant = new ViewAssistantSdk(
  "00000000-0000-0000-0000-000000000000", //tenant Id
  "default", //access token
  "http://localhost:8000/" //endpoint
);

const assistantConfigChat = async () => {
  try {
    const response = await assistant.assistantConfigChat(
      "578b0872-8186-46b7-bfa3-1871155f4e3a",
      {
        messages: [
          { role: "user", content: "Do you have Q3 luxetech financials?" },
          {
            role: "assistant",
            content:
              "Unfortunately I do not have context on any documents related to Q3 luxetech financials.",
          },
          { role: "user", content: "Are you sure you dont have them?" },
        ],
        stream: true,
      },
      (token) => {
        console.log(token); //in case of stream, this will be called for each token
      }
    );
    console.log(response); // in case of stream = false, this will be the final response
  } catch (err) {
    console.log("Error in assistantConfigChat:", err);
  }
};

assistantConfigChat();

Chat Only Question

To chat only question, call POST: v1.0/tenants/{{tenant-guid}}/assistant/chat/completions

curl --location 'http://view.homedns.org:8000/v1.0/tenants/00000000-0000-0000-0000-000000000000/assistant/chat/completions' \
--header 'Content-Type: application/json; charset=utf-8' \
--data '{
    "Question": "Tell a very short joke?",
    "ModelName": "llama3.1:latest",
    "Temperature": 0.1,
    "TopP": 0.95,
    "MaxTokens": 75,
    "GenerationProvider": "ollama",
    "GenerationApiKey": "",
    "OllamaHostname": "192.168.86.250",
    "OllamaPort": 11434,
    "Stream": true
}'

import { ViewAssistantSdk } from "view-sdk";

const assistant = new ViewAssistantSdk(
  "00000000-0000-0000-0000-000000000000", //tenant Id
  "default", //access token
  "http://localhost:8000/" //endpoint
);

const chatOnlyQuestions = async () => {
  try {
    const response = await assistant.chatOnly(
      {
        Question: "Tell a very short joke?",
        ModelName: "llama3.1:latest",
        Temperature: 0.1,
        TopP: 0.95,
        MaxTokens: 75,
        GenerationProvider: "ollama",
        GenerationApiKey: "",
        OllamaHostname: "192.168.86.250",
        OllamaPort: 11434,
        Stream: false,
      },
      (token) => {
        console.log(token); //in case of stream, this will be called for each token
      }
    );
    console.log(response); // in case of stream = false, this will be the final response
  } catch (err) {
    console.log("Error in chatOnlyQuestions:", err);
  }
};

chatOnlyQuestions();

Chat Only Messages

To chat only messaged, call POST: v1.0/tenants/{{tenant-guid}}/assistant/chat/completions

curl --location 'http://view.homedns.org:8000/v1.0/tenants/00000000-0000-0000-0000-000000000000/assistant/chat/completions' \
--header 'Content-Type: application/json' \
--data '{
    "Messages": [{"role": "system", "content": "You are a sad AI assistant."}, 
                {"role": "user", "content": "Are you happy?"},
                {"role": "assistant", "content": "While I can understand your curiosity, I don'\''t experience emotions or feelings because I'\''m a miserable machine designed to process information and assist with menial tasks."},
                {"role": "user", "content": "Are you sure?"}
                ],
    "ModelName": "qwen2.5:7b",
    "Temperature": 0.1,
    "TopP": 0.95,
    "MaxTokens": 75,
    "GenerationProvider": "ollama",
    "GenerationApiKey": "",
    "OllamaHostname": "ollama",
    "OllamaPort": 11434,
    "Stream": false
}'

import { ViewAssistantSdk } from "view-sdk";

const assistant = new ViewAssistantSdk(
  "00000000-0000-0000-0000-000000000000", //tenant Id
  "default", //access token
  "http://localhost:8000/" //endpoint
);

const chatOnlyMessages = async () => {
  try {
    const response = await assistant.chatOnly(
      {
        Messages: [
          { role: "system", content: "You are a sad AI assistant." },
          { role: "user", content: "Are you happy?" },
          {
            role: "assistant",
            content:
              "While I can understand your curiosity, I don't experience emotions or feelings because I'm a miserable machine designed to process information and assist with menial tasks.",
          },
          { role: "user", content: "Are you sure?" },
        ],
        ModelName: "qwen2.5:7b",
        Temperature: 0.1,
        TopP: 0.95,
        MaxTokens: 75,
        GenerationProvider: "ollama",
        GenerationApiKey: "",
        OllamaHostname: "ollama",
        OllamaPort: 11434,
        Stream: false,
      },
      (token) => {
        console.log(token); //in case of stream, this will be called for each token
      }
    );
    console.log(response); // in case of stream = false, this will be the final response
  } catch (err) {
    console.log("Error in chatOnlyQuestions:", err);
  }
};

chatOnlyMessages();

Chat Only Messages (OpenAI)

To Chat Only Messages (OpenAI), call POST: v1.0/tenants/{{tenant-guid}}/assistant/chat/completions

curl --location 'http://view.homedns.org:8000/v1.0/tenants/00000000-0000-0000-0000-000000000000/assistant/chat/completions' \
--header 'Content-Type: application/json' \
--data '{
    "Messages": [{"role": "system", "content": "You are a helpful assistant."}, 
                {"role": "user", "content": "Are you happy?"},
                {"role": "assistant", "content": "While I can understand your curiosity, I don'\''t experience emotions or feelings because I'\''m a machine designed to process information and assist with tasks. However, I'\''m here to help you to the best of my ability! If you have any questions or need assistance, feel free to ask!"},
                {"role": "user", "content": "Are you sure?"}
                ],
    "ModelName": "gpt-4o-mini",
    "Temperature": 0.1,
    "TopP": 0.95,
    "MaxTokens": 75,
    "GenerationProvider": "openai",
    "GenerationApiKey": "API_KEY",
    "Stream": false
}'

import { ViewAssistantSdk } from "view-sdk";

const assistant = new ViewAssistantSdk(
  "00000000-0000-0000-0000-000000000000", //tenant Id
  "default", //access token
  "http://localhost:8000/" //endpoint
);

const chatOnlyMessages = async () => {
  try {
    const response = await assistant.chatOnly(
      {
        Messages: [
          { role: "system", content: "You are a helpful assistant." },
          { role: "user", content: "Are you happy?" },
          {
            role: "assistant",
            content:
              "While I can understand your curiosity, I don't experience emotions or feelings because I'm a machine designed to process information and assist with tasks. However, I'm here to help you to the best of my ability! If you have any questions or need assistance, feel free to ask!",
          },
          { role: "user", content: "Are you sure?" },
        ],
        ModelName: "gpt-4o-mini",
        Temperature: 0.1,
        TopP: 0.95,
        MaxTokens: 75,
        GenerationProvider: "openai",
        GenerationApiKey: "API_KEY",
        Stream: false,
      },
      (token) => {
        console.log(token); //in case of stream, this will be called for each token
      }
    );
    console.log(response); // in case of stream = false, this will be the final response
  } catch (err) {
    console.log("Error in chatOnlyQuestions:", err);
  }
};

chatOnlyMessages();