Flattening a List of Objects for GPT Input in Next Column

Owen E. · 2025-07-24T17:09:47.200Z

Any way of flattening a list to use in the next column? Currently have an http request that returns list of objects like {id: 2, text: text} and want to give this to the next column which calls gpt on the list of texts without the id part. I cant use an intermediary formula column because the texts are over the 8kb cell limit. Any help would be appreciated!

Flattening a List of Objects for GPT Input in Next Column | Clay

7 comments

· Sorted by Oldest

Clay
APP
·
·
You can flatten your list of objects using the "Send row for each item in a list" feature, which is specifically designed for flattening lists. Here's how: 1. Use "Send row for each item in a list" - This turns each item in your list into its own row in another table 2. Alternative method: In Cell details, hover to the right of your list and select "Turn items into rows" to move each list item into its own row in another table This approach bypasses the 8kb cell limit since you're sending data to separate rows rather than processing it within a single cell. Each row can then reference just the text portion of your objects for your GPT column.
Sources: Send table data | Documentation | Clay University Manage cell data | Documentation | Clay University Manage cell data | Documentation | Clay University
Owen E.
·
·
These are not quite the solutions Im looking for, basically looking to get this
JSON.stringify({{Reviews}}?.map(review => review?.snippet))
into the gpt text column but I dont see an option for a formula like there is on regular text boxes
Clay
APP
·
·
You can switch your GPT column to use formulas instead of the default "Text with tokens" format. To do this: 1. Click on your GPT column header 2. Select "Edit column" from the dropdown 3. Click the gear icon and select "Formulas" as your input type 4. Save your changes Once switched to formula mode, you can use JavaScript expressions like JSON.stringify({{Reviews}}?.map(review => review?.snippet)) directly in your GPT column. Alternatively, you could use our "Extract Values from Data" integration to pull specific values from your list of objects, which might be simpler than writing custom JavaScript.
Sources: Table columns | Documentation | Clay University AI Formulas + Conditional Runs [Clay 101] - Clay University Extract Values from Data with Clay
Owen E.
·
·
I think this option is missing on the chatgpt generate text column
Clay
APP
·
·
You're correct that the ChatGPT Generate Text column doesn't have a formula option like regular formula columns do. For your specific use case, you'll need to create a separate Formula column first to transform your data using JSON.stringify({{Reviews}}?.map(review => review?.snippet)), then reference that formula column's output in your ChatGPT column. Here's the workflow: 1. Add a new Formula column 2. Use the AI Formula Generator to create your JSON transformation 3. Reference this formula column in your ChatGPT Generate Text column's prompt field This approach lets you process your list transformation without hitting the 8kb cell limit, since formula columns handle data processing efficiently.
Sources: ChatGPT: Generate text with OpenAI/GPT Formula generator | Documentation | Clay University Formula generator | Documentation | Clay University Formulas | Documentation | Clay University
Owen E.
·
·
https://app.clay.com/workspaces/674423/workbooks/wb_0szuvl9ee7yJWhGiqz7/tables/t_0szvg10mDBBvwS2T8jE/views/gv_0szvg11oFpDxqPrTGmY
Hari K.
·
·
Hey Owen! Happy to help with this - which HTTP API pull are you referring to?

Clay
APP
·
·
You can flatten your list of objects using the "Send row for each item in a list" feature, which is specifically designed for flattening lists. Here's how: 1. Use "Send row for each item in a list" - This turns each item in your list into its own row in another table 2. Alternative method: In Cell details, hover to the right of your list and select "Turn items into rows" to move each list item into its own row in another table This approach bypasses the 8kb cell limit since you're sending data to separate rows rather than processing it within a single cell. Each row can then reference just the text portion of your objects for your GPT column.
Sources: Send table data | Documentation | Clay University Manage cell data | Documentation | Clay University Manage cell data | Documentation | Clay University
Owen E.
·
·
These are not quite the solutions Im looking for, basically looking to get this
JSON.stringify({{Reviews}}?.map(review => review?.snippet))
into the gpt text column but I dont see an option for a formula like there is on regular text boxes
Clay
APP
·
·
You can switch your GPT column to use formulas instead of the default "Text with tokens" format. To do this: 1. Click on your GPT column header 2. Select "Edit column" from the dropdown 3. Click the gear icon and select "Formulas" as your input type 4. Save your changes Once switched to formula mode, you can use JavaScript expressions like JSON.stringify({{Reviews}}?.map(review => review?.snippet)) directly in your GPT column. Alternatively, you could use our "Extract Values from Data" integration to pull specific values from your list of objects, which might be simpler than writing custom JavaScript.
Sources: Table columns | Documentation | Clay University AI Formulas + Conditional Runs [Clay 101] - Clay University Extract Values from Data with Clay
Owen E.
·
·
I think this option is missing on the chatgpt generate text column
Clay
APP
·
·
You're correct that the ChatGPT Generate Text column doesn't have a formula option like regular formula columns do. For your specific use case, you'll need to create a separate Formula column first to transform your data using JSON.stringify({{Reviews}}?.map(review => review?.snippet)), then reference that formula column's output in your ChatGPT column. Here's the workflow: 1. Add a new Formula column 2. Use the AI Formula Generator to create your JSON transformation 3. Reference this formula column in your ChatGPT Generate Text column's prompt field This approach lets you process your list transformation without hitting the 8kb cell limit, since formula columns handle data processing efficiently.
Sources: ChatGPT: Generate text with OpenAI/GPT Formula generator | Documentation | Clay University Formula generator | Documentation | Clay University Formulas | Documentation | Clay University
Owen E.
·
·
https://app.clay.com/workspaces/674423/workbooks/wb_0szuvl9ee7yJWhGiqz7/tables/t_0szvg10mDBBvwS2T8jE/views/gv_0szvg11oFpDxqPrTGmY
Hari K.
·
·
Hey Owen! Happy to help with this - which HTTP API pull are you referring to?