AI credits and AI packages in Creatio
Creatio provides a unified platform approach to consuming AI capabilities across both the Creatio AI CRM and custom applications or agents built on the Creatio AI platform. From personal agents that assist employees to more complex enterprise agents that execute multi-step business processes, organizations can use a common consumption model based on AI credits.
AI credits are a universal currency that standardizes how supported AI services are measured and managed. Organizations can share a credit allowance across use cases, choose appropriate models, estimate demand, and monitor usage as adoption grows.
AI credits
An AI credit is a standardized unit that measures how much of the supported AI services your organization consumes in Creatio. Credits are used each time an AI feature performs a metered operation. For example, an AI feature can do the following:
- Generate personalized sales emails and marketing content.
- Summarize customer interactions, meetings, or service cases.
- Extract and analyze information from documents.
- Answer questions using enterprise knowledge.
- Run AI-powered voice interactions and multi-step workflows.
One request does not necessarily equal one credit. An agent can call multiple models or services to complete one business task.
AI credits and tokens
Tokens are the units of content that AI models process. Input tokens include prompts, retrieved knowledge, documents, and conversation context. Output tokens are the generated responses. AI credits add one usage layer across all supported models and services. Credits are not a fixed one-to-one conversion from tokens.
AI tokens | Creatio AI credits |
|---|---|
Units of content that a given model processes. | Common usage units in Creatio. |
Input and output amounts vary by task. | Reflect the model and service usage. |
Different models have different token rates. | Let you plan and monitor in one place. |
One allowance across models and modalities
Creatio supports multiple AI model providers through one AI credit framework. You can select different supported providers and models for different tasks, and all of them draw from the same AI credit allowance. You do not need a separate credit balance for every model provider.
The same approach covers multiple AI modalities, such as web chat, voice, video, and image recognition. One allowance can cover text generation, reasoning, document and image understanding, speech-to-text, text-to-speech, and voice. What it covers depends on the supported services and your setup. Usage varies by model, modality, and the amount of processing. The allowance is shared, but the rate for each operation is not the same.
For example, a customer service agent can turn a caller's speech into text and use a language model to understand the request. Then it finds the facts it needs and answers out loud. All metered AI services in this conversation count against the same credit allowance, even when different model providers power individual steps.
The models and modalities you can use depend on your Creatio setup, the supported integrations, and your service terms.
How AI credit usage is calculated
Usage depends on the following factors:
- Model selection. Lightweight models suit routine classification, extraction, and drafting. More capable reasoning models suit complex analysis. The model you select can change the credits that similar tasks use.
- Input and output volume. Long prompts, long conversation history, large documents, and retrieved knowledge increase the processing. Output length can also affect usage. Input and output tokens can be metered differently, depending on the model.
- Multiple agent steps. An agent can classify a request, retrieve information, analyze a document, generate a response, and evaluate the result. Each metered AI operation adds to the total. Calls to external business systems do not use AI credits on their own. Only the AI services in those steps use credits.
- Additional AI services. Speech recognition, speech generation, and document or image processing can also add to usage, as can other metered services. Voice agents in particular can use multiple services during one conversation.
For example, an agent can analyze a customer request, summarize the relevant information, and generate a recommended response. The calculation below assumes GPT-4.1 mini and three model calls, each rounded to a whole credit.
AI processing step | Illustrative credits |
|---|---|
Analyze customer request | 1 |
Summarize relevant information | 1 |
Generate recommended response | 1 |
Total per interaction | 3 |
These values are hypothetical teaching examples, not published Creatio rates. Measure the actual usage in a test environment that matches your real work.
AI packages
Creatio AI packages are more than a bundle of LLM tokens. They combine the following components to help organizations adopt, operate, and scale enterprise AI:
- AI credits. A shared allowance across supported AI models, providers, modalities, AI Twins, agents, and workflows.
- Expanded AI infrastructure. More platform resources and computing power for complex agents, bigger workloads, and large-scale deployments.
Together, these components help organizations move from initial experimentation to production deployments and broader enterprise adoption.
Package | Included AI credits | Intended usage |
|---|---|---|
Grow | 1,500,000 | Initial AI adoption and production deployment of priority use cases. |
Accelerate | 4,000,000 | Multiple AI agents, expanded infrastructure, and adoption across teams. |
Scale | 8,000,000 | Enterprise-wide AI adoption, higher-volume workloads, and advanced governance. |
Additional AI credits | 500,000 | Extend capacity at any package level. |
Each package provides a combination of AI usage capacity and infrastructure entitlements that suits a different stage of AI maturity.
You can buy additional AI credits in blocks and add them to any package. This way, you increase capacity without moving to a different package, and you scale AI usage apart from your infrastructure needs.
Learn more about package pricing: Creatio pricing.
Credit rollover
AI usage changes as you add use cases, grow adoption, and adjust your business. A shared allowance and flexible usage over time make AI usage easier to predict.
Unused AI credits roll over from month to month and across subscription years within the contracted term. You do not lose unused capacity at the end of each month or subscription year. For example, an organization can consume fewer credits while it develops its first agents and significantly more after the production rollout. Its available credits support both phases, so usage does not have to follow a fixed monthly pattern.
This flexibility helps organizations plan AI subscriptions with greater confidence and adopt AI at their own pace.
Estimate AI credit requirements
A practical forecast starts with a business activity, not with the number of user licenses. Multiply the number of operations by the average credits per operation.
For complex agents, measure the complete workflow, including all its model and service calls.
The examples below are calculated planning examples, not measured Creatio production benchmarks. Each example states its reference model and the token volumes it assumes, so you can reproduce it.
For example, AI Twin can draft personalized sales emails, such as customer follow-ups based on opportunity and relationship context. Reference model: OpenAI GPT-4.1 mini (text generation). Each email assumes 3,000 input tokens and 1,000 output tokens, one metered model call, rounded up to 1 credit.
Metric | Illustrative value |
|---|---|
Sales users | 100 |
Emails per user per day | 10 |
Working days per month | 20 |
Monthly emails | 20,000 |
Average credits per email | 1 |
Estimated monthly credits | 20,000 |
Estimated annual credits | 240,000 |
Reference AI model | OpenAI GPT-4.1 mini |
As another example, an agent can summarize completed service conversations and propose next steps. Reference model: OpenAI GPT-4.1 mini (text summarization). Each summary assumes 6,000 input tokens and 500 output tokens, one metered model call, rounded up to 1 credit.
Metric | Illustrative value |
|---|---|
Service representatives | 50 |
Summaries per representative per day | 30 |
Working days per month | 20 |
Monthly summaries | 30,000 |
Average credits per summary | 1 |
Estimated monthly credits | 30,000 |
Estimated annual credits | 360,000 |
Reference AI model | OpenAI GPT-4.1 mini |
Build a reliable forecast
A reliable forecast starts with how your organization will use AI. Look at the business activities you automate, how often they run, and how complex each AI operation is. The number of licensed users alone is not enough.
- Identify priority use cases. List the AI Twins, agents, and workflows you expect to deploy, and the activities each one performs.
- Estimate transaction volumes. Determine how often each operation runs, based on users, transactions per day, and expected adoption.
- Select reference AI models. Identify the models and modalities each operation uses, because usage varies by model and service.
- Measure typical usage. Test typical scenarios to find the average credits per operation, including variations in input size, complexity, and agent steps.
- Account for adoption and growth. Plan for phased rollouts, seasonal changes, and growing usage as more teams and processes adopt AI.
- Consolidate requirements. Combine the estimates of all use cases to find your total monthly and annual AI credit requirements.
These estimates are a starting point for choosing an AI package. Refine them as real usage data becomes available.
Monitor and optimize usage
Once your agents are live, use the monitoring and observability tools of Creatio AI Studio to track actual usage and keep improving efficiency. Monitoring shows how credits are used, where actual usage differs from your estimate, and which configurations to adjust as adoption grows.
- Monitor credit usage. Track usage across agents and AI operations to see usage trends and remaining allowances.
- Compare actual usage against forecasts. Review the credits per operation and the total volumes. Look for big differences from your estimates.
- Choose models wisely. Use efficient models for routine tasks, and stronger models only where a task needs more reasoning.
- Make agents leaner. Cut extra model calls, long context, and repeated work, and keep the output quality.
- Set usage controls. Configure budgets, usage limits, and governance policies to manage usage.
- Review and adjust regularly. Update forecasts as adoption grows, new use cases appear, or existing agents change.
By combining an initial forecast with ongoing monitoring and optimization, organizations can make usage more predictable, choose the right AI package capacity, and get more business value from their AI investments.
Learn more: Usage and budgets, Set a usage budget or spending limit.
Bring your own model (BYOM)
Creatio supports a multi-model approach to AI. You can select from supported AI providers or bring your own AI models. This gives you more control over model selection, provider relationships, data governance, and your existing AI investments.
Creatio AI credits cover the following components:
- AI model access and consumption. The AI services that model vendors such as OpenAI, Anthropic, and Google provide. This includes processing tokens, generating content, reasoning, and services such as speech recognition and voice generation.
- Creatio AI platform services. What you need to deploy, run, and manage AI securely in your business. They include running agents and workflows, infrastructure and compute, security and governance, and monitoring. They also include ongoing platform updates and support.
With BYOM, you handle the AI model access part yourself. You use your own provider agreements, model endpoints, or supported AI infrastructure. BYOM does not remove the need for a Creatio AI package or AI credits. It changes which parts of AI usage you manage directly and which ones Creatio continues to deliver. Your AI credit usage reflects the Creatio platform services you use. You manage the model provider arrangements yourself. The details depend on the supported BYOM configuration and the commercial terms.
Summary
Creatio AI credits offer a unified way to measure and manage AI usage across supported models, agents, and workflows. AI packages provide credit allowances for different levels of adoption. The most reliable way to plan usage is to measure representative business transactions, forecast volumes, and monitor actual usage over time.