Significant-Gravitas/AutoGPT · error · InsufficientBalanceError

str(e)

Error message

str(e)

What it means

This is the HTTP 402 detail produced when `charge_for_direct_block_execution` raises `InsufficientBalanceError`: `str(e)` carries the billing message (typically the required credit amount vs. remaining balance). Direct external block executions are metered upfront; if the user's credit balance can't cover the block's cost, the request fails with Payment Required before execution.

Solutions

1. Top up the user's credit balance in the platform, then retry the execution.
2. Switch the block/model configuration to a cheaper option to fit the current balance.
3. Monitor balance via the platform's user/credit endpoints and alert before it hits zero for unattended automations.

Example fix

# before
POST /blocks/{llm_block_id}/execute  # 402 {"detail": "Insufficient balance: need 10, have 2"}

# after: top up credits, then retry
POST /blocks/{llm_block_id}/execute  # 200

balance = client.get("/users/me/credits").json()["balance"]
if balance < estimated_block_cost(block_id):
    raise InsufficientCredits(f"need ~{estimated_block_cost(block_id)}, have {balance}")

try:
    client.post(f"/blocks/{block_id}/execute", json=data)
except HTTPError as e:
    if e.response.status_code == 402:
        top_up_credits(min_amount=parse_needed(e.response.text))
        client.post(f"/blocks/{block_id}/execute", json=data)  # retry after top-up
    else:
        raise

When it happens

Trigger: POST `/blocks/{block_id}/execute` when the API key owner's credit balance is lower than the execution cost of the block (LLM blocks, metered API blocks).

Common situations: Exhausted free-tier credits on a hosted deployment; long-running automation draining balance mid-run; new expensive model selected on a block raising per-run cost.

Related errors

• Insufficient balance of
• Insufficient credits to reset your rate limit.
• Rate limit reset is not available (credit system is…
• str(e)
• Amount must be greater than or equal to 500

AI-assisted analysis of Significant-Gravitas/AutoGPT@9c8bb5550f (2026-08-14). Data as JSON: /api/errors/ede78343f747e992. Report an issue: GitHub.

Appendix: source

Thrown at autogpt_platform/backend/backend/api/external/v1/routes.py:117

    # consistent with chat / internal block / internal graph routes.
    await enforce_payment_paywall(auth.user_id)

    obj = backend.blocks.get_block(block_id)
    if not obj:
        raise HTTPException(status_code=404, detail=f"Block #{block_id} not found.")
    if obj.disabled:
        raise HTTPException(status_code=403, detail=f"Block #{block_id} is disabled.")

    user = await user_db.get_user_by_id(auth.user_id)
    if not user:
        raise HTTPException(status_code=404, detail="User not found.")

    try:
        await charge_for_direct_block_execution(
            user_id=auth.user_id, block=obj, input_data=data, source="external"
        )
    except InsufficientBalanceError as e:
        raise HTTPException(
            status_code=status.HTTP_402_PAYMENT_REQUIRED, detail=str(e)
        ) from e

    # Direct block execution has no graph; build a minimal ExecutionContext
    # carrying the caller's identity + timezone so blocks that depend on
    # those (e.g. time blocks) get correct data.
    execution_context = ExecutionContext(
        user_id=auth.user_id,
        user_timezone=get_user_timezone_or_utc(user.timezone),
    )

    output = defaultdict(list)
    async for name, data in obj.execute(data, execution_context=execution_context):
        output[name].append(data)
    return output


@v1_router.post(

View on GitHub (pinned to 9c8bb5550f)