Skip to main content
POST
/
v1
/
mcp
/
batch-query
Batch Query
curl --request POST \
  --url https://api.fltr.com/v1/mcp/batch-query \
  --header 'Content-Type: application/json' \
  --data '
{
  "queries": [
    {}
  ],
  "dataset_id": "<string>",
  "limit": 123,
  "rerank": true
}
'

Documentation Index

Fetch the complete documentation index at: https://docs.tryfltr.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Request

Executes multiple search queries in parallel for better performance.

​
Body

​
queries
array
required
Array of search queries (max 10)
​
dataset_id
string
required
Dataset to search
​
limit
integer
default:5
Results per query
​
rerank
boolean
default:false
Enable reranking for all queries

​
Examples

cURL
curl -X POST https://api.fltr.com/v1/mcp/batch-query \
  -H "Authorization: Bearer fltr_sk_abc123..." \
  -H "Content-Type: application/json" \
  -d '{
    "queries": [
      "How do I authenticate?",
      "What are rate limits?",
      "How to upload documents?"
    ],
    "dataset_id": "ds_abc123",
    "limit": 3
  }'
Python
response = requests.post(
    "https://api.fltr.com/v1/mcp/batch-query",
    headers={"Authorization": "Bearer fltr_sk_abc123..."},
    json={
        "queries": [
            "How do I authenticate?",
            "What are rate limits?",
            "How to upload documents?"
        ],
        "dataset_id": "ds_abc123",
        "limit": 3
    }
)

data = response.json()
for i, query_results in enumerate(data['results']):
    print(f"Query {i+1}: {len(query_results['results'])} results")

​
Response

{
  "results": [
    {
      "query": "How do I authenticate?",
      "results": [
        {
          "chunk_id": "ch_xyz789",
          "content": "...",
          "score": 0.89
        }
      ]
    },
    {
      "query": "What are rate limits?",
      "results": [...]
    },
    {
      "query": "How to upload documents?",
      "results": [...]
    }
  ],
  "query_time_ms": 120
}

​
Benefits

  • Faster: Single request vs multiple
  • Efficient: Parallel execution
  • Cost-effective: Batch counting

​
Limits

  • Max 10 queries per batch
  • Same limits as single query