Start exploring our APIs in seconds.
The API is also accessible via our Typescript SDK where we also offer a Types package for you to get return types and more to directly use in Typescript or reference in case you use Python or other languages.
If you're not authenticated, click Get API Key below. You'll be redirected to the developer console to log in or sign up. Once authenticated, your API key will appear here.
- [Authentication](https://docs.getembed.ai/reference/authentication.md): If you don't already have one, you will need to register for an account from our console, after which we automatically generate an API key for you. If the browser doesn't redirect you back to this page, come back to https://docs.mbd.xyz/reference and your API key should be available here. - [My Requests](https://docs.getembed.ai/reference/my-requests.md) - [/casts/feed/for-you](https://docs.getembed.ai/reference/post_casts-feed-for-you.md): ## Recommend casts for a specific user This API returns a list of recommended casts for a specific user (casts that a user is most likely to interact with). This API supports the `feed_id` parameter, a way to specify all filter options through the console without the need to do so in the API calls. See our [guide on feed_id](https://docs.mbd.xyz/docs/feed-builder-feed_id) for more information. If the same parameters already specified in the `feed_id` are supplied in the API call, the values in saved in the `feed_id` will be overridden. - [/casts/feed/trending](https://docs.getembed.ai/reference/post_casts-feed-trending.md): ## Get currently trending casts This API returns a list of trending casts on Farcaster. Our models detect casts that are getting more attention before they become hot and popular. This API has 2 modes of operation: 1) `top_k` (impression) mode and 2) `paging` mode. `top_k` mode is used if `top_k` is specified (and `page_size`, `page_number` parameters ignored), otherwise `paging` mode is used. 1) In `top_k` mode, you specify a `top_k` along with an `impression_count`. `impression_count` specifies the number of items in the returned results to be considered "seen" by the viewing user, and those items will be excluded in subsequent API calls. If `impression_count` is 0, the same list is returned everytime. 2) In `paging` mode, you can request for a specific range of items using a simple paging mechanism specified by `page_size` and `page_number`. No impression information is taken into consideration when returning the items list. This API supports the `feed_id` parameter, a way to specify all filter options through the console without the need to do so in the API calls. See our [guide on feed_id](https://docs.mbd.xyz/docs/feed-builder-feed_id) for more information. If the same parameters already specified in the `feed_id` are supplied in the API call, the values in saved in the `feed_id` will be overridden. - [/casts/feed/popular](https://docs.getembed.ai/reference/post_casts-feed-popular.md): ## Get a list of popular casts This API returns a list of popular casts on Farcaster. This API has 2 modes of operation: 1) `top_k` (impression) mode and 2) `paging` mode. `top_k` mode is used if `top_k` is specified (and `page_size`, `page_number` parameters ignored), otherwise `paging` mode is used. 1) In `top_k` mode, you specify a `top_k` along with an `impression_count`. `impression_count` specifies the number of items in the returned results to be considered "seen" by the viewing user, and those items will be excluded in subsequent API calls. If `impression_count` is 0, the same list is returned everytime. 2) In `paging` mode, you can request for a specific range of items using a simple paging mechanism specified by `page_size` and `page_number`. No impression information is taken into consideration when returning the items list. This API supports the `feed_id` parameter, a way to specify all filter options through the console without the need to do so in the API calls. See our [guide on feed_id](https://docs.mbd.xyz/docs/feed-builder-feed_id) for more information. If the same parameters already specified in the `feed_id` are supplied in the API call, the values in saved in the `feed_id` will be overridden. - [/casts/search/semantic](https://docs.getembed.ai/reference/post_casts-search-semantic.md): ## Semantic search on casts This API returns a list of casts on Farcaster based on the input semantic query - [/casts/labels/for-items](https://docs.getembed.ai/reference/post_casts-labels-for-items.md): ## Get AI labels for casts This API returns AI-generated labels for casts. (Please note that we support AI labels only for English content at the moment) - [/casts/labels/for-text](https://docs.getembed.ai/reference/post_casts-labels-for-text.md): ## Get AI labels for text input This API takes a list of text inputs and returns AI-generated labels for the text. (Please note that we support AI labels only for English content at the moment) - [/casts/labels/top-items](https://docs.getembed.ai/reference/post_casts-labels-top-items.md): ## Get a list of items with the highest or lowest score of an AI label This API takes a AI label name as input and returns a list of items with either the highest or lowest scores for that label (Please note that we support AI labels only for English content at the moment) - [/users/feed/similar](https://docs.getembed.ai/reference/post_users-feed-similar.md): ## Get similar users This API returns a list of similar users for an input user_id (fid) - [/users/search/semantic](https://docs.getembed.ai/reference/post_users-search-semantic.md): ## Search for users based on text This API takes text as input and searches for users that are "similar" to the content of the text - [/users/labels/for-users](https://docs.getembed.ai/reference/post_users-labels-for-users.md): ## Get AI labels for users This API returns AI-generated labels for users. - [/users/labels/top-users](https://docs.getembed.ai/reference/post_users-labels-top-users.md): ## Get a list of users with the highest or lowest score of an AI label This API takes a AI label name as input and returns a list of users with either the highest or lowest scores for that label. To provide more control for exploration, you can specify the minimum activity count required for the user to be considered, and also a `ratio` range and a `confidence` range to be considered. ratio is the percentage of the user's activity related ot the specified label, and confidence is the AI model confidence score (returned as `score` in the API results) The API returns the `user_id` (fid), `score` (for the label), `count` (number of activities related to the label), and `ratio` (percentage of activities related to the label) - [/users/feed/for-channel (sample)](https://docs.getembed.ai/reference/post_users-feed-for-channel.md): ## Get users for a channel (sample based on older data - contact us if you would like real-time support for this feature) This API returns a list of users based on a channel and event_type. - [/users/feed//for-item (sample)](https://docs.getembed.ai/reference/post_users-feed-for-item.md): ## Get users for an item (sample based on older data - contact us if you would like real-time support for this feature) This API returns a list of users based on an item_id and event_type. - [/users/feed/for-topic (sample)](https://docs.getembed.ai/reference/post_users-feed-for-topic.md): ## Get users for a topic (sample based on older data - contact us if you would like real-time support for this feature) This API returns a list of users based on an topic and event_type. - [Compute units pricing](https://docs.getembed.ai/reference/compute-units-pricing.md): Pricing in compute units for each API