Credits

Credits are the currency used on Gumloop to power your automations. Every workflow run and agent interaction consumes credits based on the resources used.

Understanding Credits

What Credits Enable

Credits let you run workflows and interact with AI agents on Gumloop. More credits = more automations.

How You Get Credits

Your subscription includes monthly credits. Purchase additional credits in bundles or enable credit overage for uninterrupted automation.

What Consumes Credits

Gumloop has two main types of automation, each with different credit models:

Workflows

Predictable costs

Run the same workflow → Pay the same credits every time

You can calculate approximate costs before running.

Agents

Variable costs

Cost depends on conversation length, AI model used, and tools called

Costs vary per interaction.

---

Workflow Credits (Predictable)

Workflows have deterministic costs - you can calculate the exact credit cost before running.

How Workflow Credits Work

Total Cost = 1 credit (base) + Node costs

Every workflow costs:

1. 1 credit to execute (base cost)
2. + Node costs based on what each node does

Node Credit Costs

Free Nodes (0 credits)

Most native Gumloop nodes cost nothing:

• Text manipulation (Combine Text, Text Formatter, Find & Replace)
• Logic (If/Else, Switch, Router)
• Loops (For Each, Loop Mode)
• Data transformation (Filter, Join, Split)
• Most integrations (Google Sheets, Slack, Gmail, Airtable, Salesforce, etc.)
• Input/Output nodes

Build workflows primarily with free nodes to minimize costs.

AI Nodes

AI nodes cost credits based on the model:

Model Type	Cost	Examples
Expert AI	30 credits	GPT 5, OpenAI o3
Advanced AI	20 credits	GPT-4.1, Claude 4 Sonnet
Standard AI	2 credits	GPT-4.1 Mini, other standard models

Provide your own API key to reduce AI node costs to 1 credit (requires Pro plan or higher).

Data & Web Nodes

Data Enrichment:

• Enrich Contact Information: 60 credits
• Enrich Company Information: 60 credits
• Search Companies: 30 credits
• Email Validator: 10 credits

Web Scraping:

• Web Agent Scraper: 10 credits
• Website Crawler: 10 credits
• Advanced Web Search: 5 credits
• Advanced Website Scraper: 2 credits
• Web Search: 2 credits
• Website Scraper: 1 credit

Custom/Code Nodes

Custom & MCP Nodes: 3 credits each

These nodes run in isolated virtual environments for security, which incurs infrastructure costs.

Example Workflow Costs

Simple Workflow (3 credits total)

Workflow: Read Google Sheet → Filter rows → Send Slack message

Step	Cost
Workflow execution	1 credit
Read Google Sheet	0 credits
Filter rows	0 credits
Send Slack message	0 credits
Total	1 credit

AI-Powered Workflow (3 credits total)

Workflow: Read emails → Categorize with AI → Update Airtable

Step	Cost
Workflow execution	1 credit
Read Gmail	0 credits
Ask AI (Standard)	2 credits
Update Airtable	0 credits
Total	3 credits

Advanced Workflow (24 credits total)

Workflow: Combine text → AI analysis → Custom processing

Step	Cost
Workflow execution	1 credit
Combine Text	0 credits
GPT-4.1 AI node	20 credits
Custom node	3 credits
Total	24 credits

Data Enrichment Workflow (122 credits total)

Workflow: Get contacts → Enrich → Update CRM

Step	Cost
Workflow execution	1 credit
Read Airtable	0 credits
Enrich Contact (×2 contacts)	120 credits
Update Salesforce	0 credits
Total	121 credits

Enrichment nodes can get expensive with loops. A workflow that enriches 100 contacts costs 6,001 credits (1 + 60×100).

Pro tip: Hover over any node’s ’?’ icon to see its credit cost.

Important Workflow Notes

Failed Workflows

If a workflow fails or stops halfway through, you’re only charged for nodes that executed before the failure.

Example: If step 3 of 5 fails, you only pay for the base cost + steps 1 and 2.

Loop Mode Costs

When nodes run in loop mode, they execute multiple times. Multiply the node cost by the number of items in your list.

Example: Enrich Contact node (60 credits) processing 10 contacts = 600 credits

---

Agent Credits (Variable)

Agents have variable costs - each conversation costs different amounts based on usage and configuration.

Three Components of Agent Costs

Every agent interaction can incur up to three types of credit charges:

AI Model Credits

Credits charged for the AI model processing your messages and deciding what actions to take. Based on token usage.

Tool Credits

Some integrations charge credits per use or per result returned. These costs are separate from AI model costs.

Workflow Credits

When agents execute workflows, the full workflow cost is added to the interaction total.

How AI Model Credits Work

AI model credits are calculated based on actual token usage. The cost depends on:

• Message length: Longer inputs and outputs consume more tokens
• Model selected: Different AI models have different per-token costs
• Conversation history: Each message includes previous context, increasing token count
• Tool definitions: More tools means more tokens in the system prompt

Important: A single user message can trigger multiple AI “steps” internally. When your agent uses tools, it goes through a reasoning loop: deciding which tool to call, processing the results, and generating a response. Each step incurs AI model credits separately, which is why tool-using conversations cost more than simple Q&A.

Model Tiers

Budget Models

Best for simple tasks and high-volume usage

Model	Best For
GPT-4.1 Mini	Simple tasks, quick responses
GPT-5 Mini	General queries, straightforward automation
Claude Haiku 4.5	Fast interactions, basic assistance

Budget models have the lowest per-token costs, making them ideal for high-volume use cases or simple tasks that don’t require advanced reasoning.

Advanced Models

Balanced performance for most use cases

Model	Best For
Claude Sonnet 4.5	Most tasks, standard operations
Claude 3.7 Sonnet	Reliable automation, balanced performance
GPT-4.1	General purpose, versatile tasks

Advanced models offer a good balance between capability and cost. They handle most business tasks effectively and are the recommended starting point for new agents.

Expert Models

Complex analysis and difficult reasoning

Model	Best For
Claude 4.1 Opus	Complex analysis, nuanced understanding
Claude 4 Opus	Difficult tasks, detailed reasoning
GPT-5	Advanced reasoning, sophisticated tasks
OpenAI o3	Complex multi-step problems

Expert models have the highest per-token costs but provide superior reasoning capabilities. Use these for tasks that truly require advanced analysis.

Tip: Start with a Budget or Advanced model and only upgrade to Expert models if needed. The actual credit cost depends on your conversation length, message complexity, and tool usage—not just the model tier.

Bring Your Own Key (BYOK): Provide your own API key to reduce agent AI model credits by 50%. See Reducing Credit Costs for details.

Typical Cost Examples

These examples illustrate how costs vary based on what your agent does. Actual costs depend on message length, conversation history, and specific tool behavior.

Scenario	Model Tier	Typical Range	What’s Happening
Simple Q&A (no tools)	Budget	1-5 credits	Single AI step, short response
Simple Q&A (no tools)	Advanced	5-15 credits	Single AI step, short response
Lookup with 1 tool call	Budget	5-15 credits	Multiple AI steps (plan, execute, respond)
Lookup with 1 tool call	Advanced	15-40 credits	Multiple AI steps (plan, execute, respond)
Research task (3+ tool calls)	Advanced	30-100 credits	Many AI steps, processing multiple results
Complex analysis	Expert	50-200+ credits	Advanced reasoning, potentially many steps
Task that runs a workflow	Any	Varies widely	AI credits + full workflow execution cost

These are rough estimates only. Your actual costs will vary based on message length, conversation history, tool result counts, and whether the agent needs multiple attempts. Run a few representative tasks and note the credit costs shown in the UI to calibrate your expectations for your specific use case.

When Agents Call Workflows

When an agent runs a workflow, you pay both costs:

Total = Agent AI credits + Workflow credits

The workflow cost is deterministic (see Workflow Credits above), but the agent’s AI credits vary based on how much reasoning is needed to decide to run the workflow and process its results.

Agents running expensive workflows can consume significant credits. Monitor usage and optimize workflows to use free nodes when possible.

---

Reducing Credit Costs

For Workflows

• Use free native nodes when possible
• Provide your own API keys (reduces AI costs to 1 credit)
• Batch operations to reduce workflow runs
• Avoid unnecessary enrichment lookups

For Agents

• Provide your own API key (50% off AI model credits)
• Choose appropriate models (don’t use Expert for simple tasks)
• Start new conversations for different topics
• Limit tools to only what’s needed
• Write clear, specific prompts
• Optimize workflows that agents call

Bring Your Own API Key (BYOK)

Provide your own API keys for AI services to significantly reduce credit costs. The discount depends on whether you’re using workflows or agents:

Automation Type	Without BYOK	With BYOK	Savings
Workflow AI Nodes	2-30 credits per call (varies by model tier)	1 credit per call	Up to 97%
Agent AI Model Credits	Full token-based cost	50% off token-based cost	50%

Tool credits and workflow base costs are not affected by BYOK — the discount applies only to AI model usage.

Requirements:

• Pro plan or higher
• Your own OpenAI, Anthropic, or Google AI account

Setup: Add your API key in Credentials Settings

---

Tracking Credit Usage

Gumloop provides a full credit usage dashboard with charts, detailed transaction logs, and export capabilities.

1. Credit Usage Dashboard

   View a daily usage chart and browse every credit transaction in your account:

   Settings > Profile > Usage & Limits

   The dashboard includes a bar chart of daily spending, period selection (billing cycle, month, year, or rolling windows), and a searchable logs table.

2. Organization Usage & Limits

   Track credit usage across your organization, view usage breakdowns, and manage credit limits:

   Organization Limits

3. Credit Logs Table

   The logs table below the chart shows every credit event. Use Grouped View (default) to see totals per workflow run or agent chat, or toggle Detailed View to see individual charges like AI node costs and tool calls.

   Filter by category (Workflow Runs, Agent Chats, etc.) and click any row to expand its breakdown. Many entries include a View link to jump directly to the source workflow run or agent conversation.

4. Organization Usage

   Organization admins and managers can view credit logs across all users from:

   Settings > Organization > Usage & Limits

   The organization view adds a User column so you can see who consumed credits.

5. Export Credit Logs

   Export credit log data as CSV for external analysis or compliance. Go to Settings > Organization > Data Export and select Credit Logs as the export type.

---

Getting More Credits

Monthly Allocation

Credits refresh monthly based on your subscription tier. See pricing page for tier details.

Credit Overage

Enable overage to continue automations after monthly credits run out.

• Cost: $0.005 per overage credit
• Limit: 2x your monthly allocation (prevents runaway costs)
• Setup: Enable in account settings

Example: 10,000 monthly credits → 20,000 max overage credits

Need More Credits?

• Temporary spike: Upgrade temporarily, then downgrade
• Consistent high usage: Contact [e-mail] for higher limits
• Bundles: Purchase credit bundles at discounted rates

Credits don’t roll over month-to-month (except Enterprise plans)

---

Why Credits?

Different automations have vastly different costs:

Action	Our Cost	Credit Charge
Text manipulation	~$0.000001	0 credits
Standard AI call	~$0.005	2 credits
Advanced AI call	~$0.05	20 credits
Contact enrichment	~$0.15	60 credits
Custom node environment	~$0.0075	3 credits

The costs above are arbitrary, this is just an example

A flat “per-execution” model would either:

• Overcharge users with lightweight workflows
• Undercharge users with expensive operations
• Force us to price for worst-case scenarios

Credits let us charge fairly based on actual resource usage.

---

Learn More

Agent Credit Details

Deep dive into agent credit costs and optimization

Pricing Plans

Compare subscription tiers and credit bundles

Credits Blog Post

Why we transitioned to credits