Credits

credits have now replaced operations as the term for make's billing unit here's what to know your plan your existing plan and pricing, including the cost of buying credits, remain unchanged the cost of credits varies by plan credit usage for non ai features your credit usage won't increase—1 operation still uses 1 credit credit usage for make's built in ai features your credit usage may increase if you're using make’s ai provider, as we align billing with actual ai usage costs this affects only certain features , such as make ai content extractor and make ai toolkit credit usage for all other ai features ai apps like openai, chatgpt, claude, etc still use 1 credit per operation the information in this article is subject to change until the credit system and its affected elements are finalized with this guide to make's credit system, you'll understand what credits are and the different types of credit usage what are credits? credits are the currency you buy and consume to use features in make your credit usage is based on the number of operations, ai tokens, and other usage based factors only features activated by scenario runs (apps, modules, and some in app features) or the ai agent’s chat feature use credits you can see credits and operations as separate terms in make because they serve different purposes credits help you track how much you’re spending, and operations docid\ zr2rwmilmj 19r85mlawm show you the outcome of your activities on the platform in make, features use credits on a fixed or dynamic basis credits /#fixed credit usage features use a set number of credits per run credits /#dynamic credit usage some ai features and advanced modules use a varying number of credits per run based on actual consumption whether a feature has fixed or dynamic credit usage depends on its processing complexity and if it uses make's ai provider in the product, make always indicates when features use credits dynamically with tags and tooltips explaining the tags in the example below, the modules use credits based on tokens fixed credit usage with fixed credit usage, modules consume credits based on a set rate by default, 1 operation uses 1 credit that rate is constant regardless of input or data transfer size most apps in make consume credits on a fixed basis, allowing for predictable costs for example, running the google drive > upload a file module always uses 1 credit per file upload, regardless of the file size exception some modules with fixed credit usage consume more than 1 credit for an operation due to their processing complexity currently, these include some make ai content extractor modules, which each use a different number of credits per operation you can identify features using more than the default 1 credit by their unique tag in the example below, the modules consume 2 or 10 credits per operation you can find detailed examples of fixed credit usage in how features use credits docid\ lzasgxrjtbcjok6vf0rkr dynamic credit usage with dynamic credit usage, modules consume a different number of credits based on actual usage rather than a fixed rate for some of make's built in ai features, you can use make's ai provider, eliminating the need to create your own account with openai, anthropic, and other providers features using make's ai provider have dynamic credit usage because make pays the provider directly since ai providers charge based on tokens, that variable cost is reflected in your credit usage most of these features consume credits based on token consumption other features, such as some make ai content extractor modules, use a different number of credits per run depending on factors like file size, page number, or run time you can identify features with dynamic credit usage by their tags, which include tokens credits based on ai tokens and/or make ai agents app docid\ mdyjgeeiirrxmd607v1xg consumed (and operations how features use credits docid\ lzasgxrjtbcjok6vf0rkr , such as with make ai agent and make ai toolkit) file size credits based on file size processed per page credits based on pages processed run time credits based on processing time for a list of all dynamic credit usage features, see how features use credits docid\ lzasgxrjtbcjok6vf0rkr credits and tokens when you use ai features in make, you'll also need to understand tokens—the unit that ai providers use to measure and charge for usage you consume tokens to interact with large language models (llms), such as chatgpt or claude token consumption depends on the amount of information processed more interactions, longer prompts, larger files, and longer responses all increase your token usage in english, 1 token equals around 4 characters , and 100 tokens equal about a paragraph when using ai in make, you have two ways to pay for your token usage use make's ai provider, or use your own connection to a provider use your own connection is available to users on pro plans and above option 1 you pay make for token usage (use make's ai provider) this option is available to users on all plans make handles the connection to an ai provider for you, so you don't need to create an account with openai, anthropic, or other providers you pay for tokens directly with make using credits you choose a large language model (llm) tier small, medium, or large each tier is a different llm that is ideal for a specific purpose llm tier token to credit usage best for small 5000 tokens per credit 5000 tokens per credit simple text analysis and basic prompts medium 3 3 500 500 tokens per credit more complex tasks and moderate file processing large 1 500 tokens per credit 500 tokens per credit for tasks requiring advanced reasoning with how features use credits docid\ lzasgxrjtbcjok6vf0rkr , you use 1 credit for each operation in addition to the token based credits above the llms used for make's ai provider and their conversion rates are subject to change based on the availability of newer models when your connection type is make's ai provider, you can see the llms behind each tier in the model dropdown in the module settings you can see the number of tokens used in each run in the white bubble that appears above a relevant module, once you expand an operation you can see both input and output tokens used option 2 you pay your ai provider for token usage (use your own connection) this option is available to users on pro plans and above you use your own connection (e g , via an api key) to an existing account with openai, anthropic, or another ai provider you pay for tokens directly with that provider you have two billing components credits make bills you for operations—1 credit per operation tokens your ai provider bills you for tokens used you pay your external ai provider based on the number of tokens consumed features offering both options if you're on a pro plan or above, you can choose between make's ai provider and your own connection for these features make ai tools docid\ qziwfqxhlmorepv5b4uqw make ai agents docid 0kbnuoduk1emvkqnwmdqc examples to understand how dynamic credit usage works in practice, consider these two examples of credit usage in make ai toolkit , an ai app that uses credits based on both operations and token consumption example 1 shorter text you use a make ai toolkit > analyze sentiment module for an email of under 500 characters, or about 60 words in a single run, the module uses 78 input tokens and 67 output tokens based on openai's token conversion rate the 145 total tokens convert to credits based on your llm tier each tier has a different conversion rate (e g , the small tier converts 5000 tokens to 1 credit) small 0 029 credits (145/5000) medium 0 04 credits (145/3500) large 0 10 credits (145/1500) total credit usage per run = token based credits + 1 credit for an operation for example, if you chose the small tier, you used 1 03 total credits (0 03 + 1 00) example 2 longer text you use a make ai toolkit > summarize sentiment module for an article of over 10,000 characters, or about 1,500 words the module uses 2170 input tokens and 145 output tokens, totalling 2315 tokens this total uses the following token based credits small 0 46 credits (2315/5000) medium 0 66 credits (2315/3500) large 1 54 (2315/1500) total credit usage per run = token based credits + 1 credit for an operation if you chose the small tier, you used 1 46 total credits (0 46 + 1 00) buying extra credits if you run out of credits before the next billing cycle starts, you can upgrade your subscription, buy extra credits, or enable extra credits auto purchasing to learn more about these options, see extra credits docid 2zcwidbk uamtp2x2 7kj the cost of buying credits remains the same as the cost of buying operations