RAG-Powered Conversational Agent

The R2R Agent endpoint provides a conversational RAG interface for retrieval, reasoning, and research. It supports multi-turn context, tool integration, streaming output, and both RAG and Research modes for dynamic AI-driven analysis.

POST /api/r2r/v3/retrieval/agent

RAG-powered Conversational Agent

cURL

curl --request POST \
  --url https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "search_settings": {
    "use_hybrid_search": false,
    "use_semantic_search": true,
    "use_fulltext_search": false,
    "filters": "<string>",
    "limit": 10,
    "offset": "0",
    "include_metadatas": true,
    "include_scores": true,
    "search_strategy": "vanilla",
    "hybrid_settings": {
      "full_text_weight": 1,
      "semantic_weight": 5,
      "full_text_limit": 200,
      "rrf_k": 50
    },
    "chunk_settings": {
      "probes": 10,
      "ef_search": 40,
      "enabled": true
    },
    "graph_settings": {
      "limits": "<array>",
      "enabled": true
    },
    "num_sub_queries": 5
  },
  "rag_generation_config": {
    "model": "<string>",
    "temperature": 123,
    "top_p": 123,
    "max_tokens_to_sample": 123,
    "stream": true,
    "functions": "<array>",
    "tools": "<array>",
    "add_generation_kwargs": "<array>",
    "api_base": "<string>",
    "response_format": [
      {
        "Base Model": {}
      }
    ],
    "extended_thinking": false,
    "thinking_budget": 123,
    "reasoning_effort": "<string>"
  },
  "research_generation_config": {
    "model": "<string>",
    "temperature": 123,
    "top_p": 123,
    "max_tokens_to_sample": 123,
    "stream": true,
    "functions": "<array>",
    "tools": "<array>",
    "add_generation_kwargs": "<array>",
    "api_base": "<string>",
    "response_format": [
      {
        "Base Model": {}
      }
    ],
    "extended_thinking": false,
    "thinking_budget": 123,
    "reasoning_effort": "<string>"
  }
}
'

import requests

url = "https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent"

payload = {
    "search_settings": {
        "use_hybrid_search": False,
        "use_semantic_search": True,
        "use_fulltext_search": False,
        "filters": "<string>",
        "limit": 10,
        "offset": "0",
        "include_metadatas": True,
        "include_scores": True,
        "search_strategy": "vanilla",
        "hybrid_settings": {
            "full_text_weight": 1,
            "semantic_weight": 5,
            "full_text_limit": 200,
            "rrf_k": 50
        },
        "chunk_settings": {
            "probes": 10,
            "ef_search": 40,
            "enabled": True
        },
        "graph_settings": {
            "limits": "<array>",
            "enabled": True
        },
        "num_sub_queries": 5
    },
    "rag_generation_config": {
        "model": "<string>",
        "temperature": 123,
        "top_p": 123,
        "max_tokens_to_sample": 123,
        "stream": True,
        "functions": "<array>",
        "tools": "<array>",
        "add_generation_kwargs": "<array>",
        "api_base": "<string>",
        "response_format": [{ "Base Model": {} }],
        "extended_thinking": False,
        "thinking_budget": 123,
        "reasoning_effort": "<string>"
    },
    "research_generation_config": {
        "model": "<string>",
        "temperature": 123,
        "top_p": 123,
        "max_tokens_to_sample": 123,
        "stream": True,
        "functions": "<array>",
        "tools": "<array>",
        "add_generation_kwargs": "<array>",
        "api_base": "<string>",
        "response_format": [{ "Base Model": {} }],
        "extended_thinking": False,
        "thinking_budget": 123,
        "reasoning_effort": "<string>"
    }
}
headers = {
    "Authorization": "Bearer <token>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    search_settings: {
      use_hybrid_search: false,
      use_semantic_search: true,
      use_fulltext_search: false,
      filters: '<string>',
      limit: 10,
      offset: '0',
      include_metadatas: true,
      include_scores: true,
      search_strategy: 'vanilla',
      hybrid_settings: {full_text_weight: 1, semantic_weight: 5, full_text_limit: 200, rrf_k: 50},
      chunk_settings: {probes: 10, ef_search: 40, enabled: true},
      graph_settings: {limits: '<array>', enabled: true},
      num_sub_queries: 5
    },
    rag_generation_config: {
      model: '<string>',
      temperature: 123,
      top_p: 123,
      max_tokens_to_sample: 123,
      stream: true,
      functions: '<array>',
      tools: '<array>',
      add_generation_kwargs: '<array>',
      api_base: '<string>',
      response_format: [{'Base Model': {}}],
      extended_thinking: false,
      thinking_budget: 123,
      reasoning_effort: '<string>'
    },
    research_generation_config: {
      model: '<string>',
      temperature: 123,
      top_p: 123,
      max_tokens_to_sample: 123,
      stream: true,
      functions: '<array>',
      tools: '<array>',
      add_generation_kwargs: '<array>',
      api_base: '<string>',
      response_format: [{'Base Model': {}}],
      extended_thinking: false,
      thinking_budget: 123,
      reasoning_effort: '<string>'
    }
  })
};

fetch('https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'search_settings' => [
        'use_hybrid_search' => false,
        'use_semantic_search' => true,
        'use_fulltext_search' => false,
        'filters' => '<string>',
        'limit' => 10,
        'offset' => '0',
        'include_metadatas' => true,
        'include_scores' => true,
        'search_strategy' => 'vanilla',
        'hybrid_settings' => [
                'full_text_weight' => 1,
                'semantic_weight' => 5,
                'full_text_limit' => 200,
                'rrf_k' => 50
        ],
        'chunk_settings' => [
                'probes' => 10,
                'ef_search' => 40,
                'enabled' => true
        ],
        'graph_settings' => [
                'limits' => '<array>',
                'enabled' => true
        ],
        'num_sub_queries' => 5
    ],
    'rag_generation_config' => [
        'model' => '<string>',
        'temperature' => 123,
        'top_p' => 123,
        'max_tokens_to_sample' => 123,
        'stream' => true,
        'functions' => '<array>',
        'tools' => '<array>',
        'add_generation_kwargs' => '<array>',
        'api_base' => '<string>',
        'response_format' => [
                [
                                'Base Model' => [
                                                                
                                ]
                ]
        ],
        'extended_thinking' => false,
        'thinking_budget' => 123,
        'reasoning_effort' => '<string>'
    ],
    'research_generation_config' => [
        'model' => '<string>',
        'temperature' => 123,
        'top_p' => 123,
        'max_tokens_to_sample' => 123,
        'stream' => true,
        'functions' => '<array>',
        'tools' => '<array>',
        'add_generation_kwargs' => '<array>',
        'api_base' => '<string>',
        'response_format' => [
                [
                                'Base Model' => [
                                                                
                                ]
                ]
        ],
        'extended_thinking' => false,
        'thinking_budget' => 123,
        'reasoning_effort' => '<string>'
    ]
  ]),
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>",
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent"

	payload := strings.NewReader("{\n  \"search_settings\": {\n    \"use_hybrid_search\": false,\n    \"use_semantic_search\": true,\n    \"use_fulltext_search\": false,\n    \"filters\": \"<string>\",\n    \"limit\": 10,\n    \"offset\": \"0\",\n    \"include_metadatas\": true,\n    \"include_scores\": true,\n    \"search_strategy\": \"vanilla\",\n    \"hybrid_settings\": {\n      \"full_text_weight\": 1,\n      \"semantic_weight\": 5,\n      \"full_text_limit\": 200,\n      \"rrf_k\": 50\n    },\n    \"chunk_settings\": {\n      \"probes\": 10,\n      \"ef_search\": 40,\n      \"enabled\": true\n    },\n    \"graph_settings\": {\n      \"limits\": \"<array>\",\n      \"enabled\": true\n    },\n    \"num_sub_queries\": 5\n  },\n  \"rag_generation_config\": {\n    \"model\": \"<string>\",\n    \"temperature\": 123,\n    \"top_p\": 123,\n    \"max_tokens_to_sample\": 123,\n    \"stream\": true,\n    \"functions\": \"<array>\",\n    \"tools\": \"<array>\",\n    \"add_generation_kwargs\": \"<array>\",\n    \"api_base\": \"<string>\",\n    \"response_format\": [\n      {\n        \"Base Model\": {}\n      }\n    ],\n    \"extended_thinking\": false,\n    \"thinking_budget\": 123,\n    \"reasoning_effort\": \"<string>\"\n  },\n  \"research_generation_config\": {\n    \"model\": \"<string>\",\n    \"temperature\": 123,\n    \"top_p\": 123,\n    \"max_tokens_to_sample\": 123,\n    \"stream\": true,\n    \"functions\": \"<array>\",\n    \"tools\": \"<array>\",\n    \"add_generation_kwargs\": \"<array>\",\n    \"api_base\": \"<string>\",\n    \"response_format\": [\n      {\n        \"Base Model\": {}\n      }\n    ],\n    \"extended_thinking\": false,\n    \"thinking_budget\": 123,\n    \"reasoning_effort\": \"<string>\"\n  }\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Authorization", "Bearer <token>")
	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent")
  .header("Authorization", "Bearer <token>")
  .header("Content-Type", "application/json")
  .body("{\n  \"search_settings\": {\n    \"use_hybrid_search\": false,\n    \"use_semantic_search\": true,\n    \"use_fulltext_search\": false,\n    \"filters\": \"<string>\",\n    \"limit\": 10,\n    \"offset\": \"0\",\n    \"include_metadatas\": true,\n    \"include_scores\": true,\n    \"search_strategy\": \"vanilla\",\n    \"hybrid_settings\": {\n      \"full_text_weight\": 1,\n      \"semantic_weight\": 5,\n      \"full_text_limit\": 200,\n      \"rrf_k\": 50\n    },\n    \"chunk_settings\": {\n      \"probes\": 10,\n      \"ef_search\": 40,\n      \"enabled\": true\n    },\n    \"graph_settings\": {\n      \"limits\": \"<array>\",\n      \"enabled\": true\n    },\n    \"num_sub_queries\": 5\n  },\n  \"rag_generation_config\": {\n    \"model\": \"<string>\",\n    \"temperature\": 123,\n    \"top_p\": 123,\n    \"max_tokens_to_sample\": 123,\n    \"stream\": true,\n    \"functions\": \"<array>\",\n    \"tools\": \"<array>\",\n    \"add_generation_kwargs\": \"<array>\",\n    \"api_base\": \"<string>\",\n    \"response_format\": [\n      {\n        \"Base Model\": {}\n      }\n    ],\n    \"extended_thinking\": false,\n    \"thinking_budget\": 123,\n    \"reasoning_effort\": \"<string>\"\n  },\n  \"research_generation_config\": {\n    \"model\": \"<string>\",\n    \"temperature\": 123,\n    \"top_p\": 123,\n    \"max_tokens_to_sample\": 123,\n    \"stream\": true,\n    \"functions\": \"<array>\",\n    \"tools\": \"<array>\",\n    \"add_generation_kwargs\": \"<array>\",\n    \"api_base\": \"<string>\",\n    \"response_format\": [\n      {\n        \"Base Model\": {}\n      }\n    ],\n    \"extended_thinking\": false,\n    \"thinking_budget\": 123,\n    \"reasoning_effort\": \"<string>\"\n  }\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.intelligence.io.solutions/api/r2r/v3/retrieval/agent")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Authorization"] = 'Bearer <token>'
request["Content-Type"] = 'application/json'
request.body = "{\n  \"search_settings\": {\n    \"use_hybrid_search\": false,\n    \"use_semantic_search\": true,\n    \"use_fulltext_search\": false,\n    \"filters\": \"<string>\",\n    \"limit\": 10,\n    \"offset\": \"0\",\n    \"include_metadatas\": true,\n    \"include_scores\": true,\n    \"search_strategy\": \"vanilla\",\n    \"hybrid_settings\": {\n      \"full_text_weight\": 1,\n      \"semantic_weight\": 5,\n      \"full_text_limit\": 200,\n      \"rrf_k\": 50\n    },\n    \"chunk_settings\": {\n      \"probes\": 10,\n      \"ef_search\": 40,\n      \"enabled\": true\n    },\n    \"graph_settings\": {\n      \"limits\": \"<array>\",\n      \"enabled\": true\n    },\n    \"num_sub_queries\": 5\n  },\n  \"rag_generation_config\": {\n    \"model\": \"<string>\",\n    \"temperature\": 123,\n    \"top_p\": 123,\n    \"max_tokens_to_sample\": 123,\n    \"stream\": true,\n    \"functions\": \"<array>\",\n    \"tools\": \"<array>\",\n    \"add_generation_kwargs\": \"<array>\",\n    \"api_base\": \"<string>\",\n    \"response_format\": [\n      {\n        \"Base Model\": {}\n      }\n    ],\n    \"extended_thinking\": false,\n    \"thinking_budget\": 123,\n    \"reasoning_effort\": \"<string>\"\n  },\n  \"research_generation_config\": {\n    \"model\": \"<string>\",\n    \"temperature\": 123,\n    \"top_p\": 123,\n    \"max_tokens_to_sample\": 123,\n    \"stream\": true,\n    \"functions\": \"<array>\",\n    \"tools\": \"<array>\",\n    \"add_generation_kwargs\": \"<array>\",\n    \"api_base\": \"<string>\",\n    \"response_format\": [\n      {\n        \"Base Model\": {}\n      }\n    ],\n    \"extended_thinking\": false,\n    \"thinking_budget\": 123,\n    \"reasoning_effort\": \"<string>\"\n  }\n}"

response = http.request(request)
puts response.read_body

200

{
  "results": {
    "messages": [
      {
        "role": "assistant",
        "content": "Aristotle (384–322 BC) was an Ancient\n                        Greek philosopher and polymath whose contributions\n                        have had a profound impact on various fields of\n                        knowledge.\n                        Here are some key points about his life and work:\n                        \n\n1. **Early Life**: Aristotle was born in 384 BC in\n                        Stagira, Chalcidice, which is near modern-day\n                        Thessaloniki, Greece. His father, Nicomachus, was the\n                        personal physician to King Amyntas of Macedon, which\n                        exposed Aristotle to medical and biological knowledge\n                        from a young age [C].\n\n2. **Education and Career**:\n                        After the death of his parents, Aristotle was sent to\n                        Athens to study at Plato's Academy, where he remained\n                        for about 20 years. After Plato's death, Aristotle\n                        left Athens and eventually became the tutor of\n                        Alexander the Great [C].\n                        \n\n3. **Philosophical Contributions**: Aristotle\n                        founded the Lyceum in Athens, where he established the\n                        Peripatetic school of philosophy. His works cover a\n                        wide range of subjects, including metaphysics, ethics,\n                        politics, logic, biology, and aesthetics. His writings\n                        laid the groundwork for many modern scientific and\n                        philosophical inquiries [A].\n\n4. **Legacy**:\n                        Aristotle's influence extends beyond philosophy to the\n                          natural sciences, linguistics, economics, and\n                          psychology. His method of systematic observation and\n                          analysis has been foundational to the development of\n                          modern science [A].\n\nAristotle's comprehensive\n                          approach to knowledge and his systematic methodology\n                          have earned him a lasting legacy as one of the\n                          greatest philosophers of all time.\n\nSources:\n                          \n- [A] Aristotle's broad range of writings and\n                          influence on modern science.\n- [C] Details about\n                          Aristotle's early life and education.",
        "metadata": {
          "aggregated_search_results": {
            "chunk_search_results": [
              {
                "document_id": "3e157b3a-8469-51db-90d9-52e7d896b49b",
                "id": "3f3d47f3-8baf-58eb-8bc2-0171fb1c6e09",
                "metadata": {
                  "associated_query": "What is the capital of France?",
                  "title": "example_document.pdf"
                },
                "owner_id": "2acb499e-8428-543b-bd85-0d9098718220",
                "score": 0.23943702876567796,
                "text": "Example text from the document"
              }
            ],
            "document_search_results": [
              {
                "document": {
                  "chunks": [
                    "Chunk 1",
                    "Chunk 2"
                  ],
                  "id": "3f3d47f3-8baf-58eb-8bc2-0171fb1c6e09",
                  "metadata": {},
                  "title": "Document Title"
                }
              }
            ],
            "graph_search_results": [
              {
                "chunk_ids": [
                  "c68dc72e-fc23-5452-8f49-d7bd46088a96"
                ],
                "content": {
                  "description": "Entity Description",
                  "id": "3f3d47f3-8baf-58eb-8bc2-0171fb1c6e09",
                  "metadata": {},
                  "name": "Entity Name"
                },
                "metadata": {
                  "associated_query": "What is the capital of France?"
                },
                "result_type": "entity"
              }
            ],
            "web_search_results": [
              {
                "date": "2021-01-01",
                "link": "https://example.com/page",
                "position": 1,
                "sitelinks": [
                  {
                    "link": "https://example.com/sitelink",
                    "title": "Sitelink Title"
                  }
                ],
                "snippet": "Page snippet",
                "title": "Page Title"
              }
            ]
          },
          "citations": [
            {
              "collection_ids": [
                "122fdf6a-e116-546b-a8f6-e4cb2e2c0a09"
              ],
              "document_id": "\n                                    e43864f5-a36f-548e-aacd-6f8d48b30c7f\n                                    ",
              "endIndex": 396,
              "id": "e760bb76-1c6e-52eb-910d-0ce5b567011b",
              "index": 1,
              "metadata": {
                "chunk_order": 68,
                "document_type": "pdf",
                "license": "CC-BY-4.0",
                "title": "DeepSeek_R1.pdf"
              },
              "owner_id": "\n                                    2acb499e-8428-543b-bd85-0d9098718220\n                                    ",
              "rawIndex": 9,
              "score": 0.64,
              "snippetEndIndex": 418,
              "snippetStartIndex": 320,
              "sourceType": "chunk",
              "startIndex": 393,
              "text": "\n                                    Document Title: DeepSeek_R1.pdf\n                                    \n\nText: could achieve an accuracy of ...\n                                    "
            }
          ]
        }
      }
    ],
    "conversation_id": "a32b4c5d-6e7f-8a9b-0c1d-2e3f4a5b6c7d"
  }
}

{}

{}

The RAG-powered Conversational Agent enables interactive, multi-turn communication with an intelligent agent built on R2R’s Retrieval-Augmented Generation (RAG) system.

This endpoint allows users to engage in real-time dialogue with an AI capable of retrieving information from internal and external sources, reasoning through complex problems, executing computations, and maintaining context across multiple conversation turns.

It operates in two distinct modes: RAG Mode for knowledge-based responses and Research Mode for deep analytical reasoning.

Operating Modes

RAG Mode (Default)

Provides fast, grounded answers by combining retrieval and generation.

Features include:

• Semantic and hybrid search across documents and chunks
• Optional web search integration for live context
• Document-level and chunk-level content retrieval
• Source citation and evidence-based responses

Research Mode

Extends RAG functionality with advanced reasoning and computational abilities.

Features include:

• Dedicated reasoning system for multi-step problem-solving
• Automated critique generation to identify biases or logical fallacies
• Python execution for quantitative analysis and code-based reasoning
• Deep exploration capabilities across multiple sources

Available Tools

RAG Tools:

• search_file_knowledge — Perform semantic or hybrid search across ingested documents.
• search_file_descriptions — Search file-level metadata and descriptions.
• content — Retrieve full documents or chunk structures.
• web_search — Query external search engines for up-to-date information.
• web_scrape — Extract content directly from specified web pages.

Research Tools:

• rag — Invoke the underlying RAG agent for information retrieval.
• reasoning — Use a dedicated reasoning model for deep analysis and logical inference.
• critique — Analyze the conversation for potential biases or reasoning flaws.
• python_executor — Execute Python code for computation, simulation, or data processing.

Streaming Output

When streaming is enabled ("stream": true), the API emits Server-Sent Events (SSE) to deliver updates in real time.

Each event corresponds to a stage in the agent’s reasoning and response generation process.

Event Type	Description
thinking	Displays the model’s intermediate reasoning steps (enabled by extended_thinking=true).
tool_call	Indicates when the agent invokes a tool.
tool_result	Contains the output from an executed tool.
citation	Signals that a citation has been added to the response.
message	Streams partial tokens of the generated message.
final_answer	Provides the complete generated response with structured citations.

Conversations

The agent maintains persistent conversational context using the conversation_id field.

How it works:

1. On the initial request, the system creates a new conversation and returns a conversation_id.
2. Include this ID in subsequent requests to continue the same thread.
3. If no conversation name exists, R2R automatically assigns one.

This design allows for multi-turn, context-aware discussions, where the agent can recall prior messages, reasoning, and results.

Authorizations

Authorization · string · header · required

The access token received from the authorization server in the OAuth 2.0 flow.

Body

application/json

Params · object

Current message to process

Show child attributes

search_mode · enum<string>

Pre-configured search modes: basic, advanced, or custom.

Available options: basic, advanced, custom

search_settings · object

The search configuration object for retrieving context.

Show child attributes

rag_generation_config · object

Configuration for RAG generation in 'rag' mode

Show child attributes

research_generation_config · object

Configuration for generation in ‘research’ mode. If not provided but mode=‘research’, rag_generation_config will be used with appropriate model overrides.

Show child attributes

rag_tools · enum<string>

List of tools to enable for RAG mode. Available tools: search_file_knowledge, get_file_content, web_search, web_scrape, search_file_descriptions

Available options: web_search, web_scrape, search_file_descriptions, search_file_knowledge, get_file_content

Response

200

application/json

200

results · object

Show child attributes