Batch transfers

When you need to send to multiple recipients, submit independent transfers with independent request_id values. CAW evaluates each transfer separately through the policy engine, so one denial does not invalidate the others.

Parallel batch transfer

CLI
Python SDK
TypeScript SDK

PACT_ID=<pact-id>
RECIPIENTS=("0xAddr1" "0xAddr2" "0xAddr3")
AMOUNTS=("10" "15" "20")

for i in "${!RECIPIENTS[@]}"; do
  caw tx transfer \
    --pact-id "$PACT_ID" \
    --dst-address "${RECIPIENTS[$i]}" \
    --token-id SETH_USDC \
    --amount "${AMOUNTS[$i]}" \
    --chain-id SETH \
    --request-id "batch-$(date +%s)-$i"
done

import asyncio

transfers = [
    {"dst_addr": "0xAddr1...", "amount": "10", "request_id": "batch-001"},
    {"dst_addr": "0xAddr2...", "amount": "15", "request_id": "batch-002"},
    {"dst_addr": "0xAddr3...", "amount": "20", "request_id": "batch-003"},
]

results = await asyncio.gather(
    *[
        client.transfer_tokens(
            wallet_uuid=wallet_id,
            chain_id="SETH",
            dst_addr=item["dst_addr"],
            token_id="SETH_USDC",
            amount=item["amount"],
            request_id=item["request_id"],
        )
        for item in transfers
    ],
    return_exceptions=True,
)

const transfers = [
  { dst_addr: '0xAddr1...', amount: '10', request_id: 'batch-001' },
  { dst_addr: '0xAddr2...', amount: '15', request_id: 'batch-002' },
  { dst_addr: '0xAddr3...', amount: '20', request_id: 'batch-003' },
];

const results = await Promise.allSettled(
  transfers.map((item) =>
    txApi.transferTokens(walletId, {
      chain_id: 'SETH',
      dst_addr: item.dst_addr,
      token_id: 'SETH_USDC',
      amount: item.amount,
      request_id: item.request_id,
    })
  )
);

Recommended pattern

Use one business-stable request_id per recipient or payout unit.
Track each result by request_id, not by array position.
Expect mixed outcomes in the same batch: allowed, pending approval, denied.

For durable tracking after submission, use Transaction Tracking.

Start here

Frameworks & MCP

Concepts

How-to

Parallel batch transfer

Recommended pattern

​Parallel batch transfer

​Recommended pattern

Parallel batch transfer

Recommended pattern