Summarize a transcript using LeMUR | AssemblyAI

1 curl -X POST https://api.assemblyai.com/lemur/v3/generate/summary \ 2 -H "Authorization: <apiKey>" \ 3 -H "Content-Type: application/json" \ 4 -d '{ 5 "final_model": "anthropic/claude-sonnet-4-20250514", 6 "context": "This is an interview about wildfires.", 7 "max_output_size": 3000, 8 "temperature": 0, 9 "transcript_ids": [ 10 "47b95ba5-8889-44d8-bc80-5de38306e582" 11 ] 12 }'

1 { 2 "request_id": "5e1b27c2-691f-4414-8bc5-f14678442f9e", 3 "response": "- Wildfires in Canada are sending smoke and air pollution across parts of the US, triggering air quality alerts from Maine to Minnesota. Concentrations of particulate matter have exceeded safety levels.\n\n- Weather systems are channeling the smoke through Pennsylvania into the Mid-Atlantic and Northeast regions. New York City has canceled outdoor activities to keep children and vulnerable groups indoors.\n\n- Very small particulate matter can enter the lungs and impact respiratory, cardiovascular and neurological health. Young children, the elderly and those with preexisting conditions are most at risk.\n\n- The conditions causing the poor air quality could get worse or shift to different areas in coming days depending on weather patterns. More wildfires may also contribute to higher concentrations.\n\n- Climate change is leading to longer and more severe fire seasons. Events of smoke traveling long distances and affecting air quality over wide areas will likely become more common in the future.\"\n", 4 "usage": { 5 "input_tokens": 27, 6 "output_tokens": 3 7 } 8 }

To use our EU server with LeMUR, replace api.assemblyai.com with api.eu.assemblyai.com.

Custom Summary allows you to distill a piece of audio into a few impactful sentences. You can give the model context to obtain more targeted results while outputting the results in a variety of formats described in human language.

<Note>To use our EU server with LeMUR, replace `api.assemblyai.com` with `api.eu.assemblyai.com`.</Note> Custom Summary allows you to distill a piece of audio into a few impactful sentences. You can give the model context to obtain more targeted results while outputting the results in a variety of formats described in human language.

Authentication

Authorizationstring

API Key authentication via header

Request

Params to generate the summary

final_modelenumRequired

The model that is used for the final prompt after compression is performed.

answer_formatstringOptional

How you want the summary to be returned. This can be any text. Examples: “TLDR”, “bullet points”

contextstring or map from strings to anyOptional

Context to provide the model. This can be a string or a free-form JSON value.

input_textstringOptional

Custom formatted transcript data. Maximum size is the context limit of the selected model. Use either transcript_ids or input_text as input into LeMUR.

max_output_sizeintegerOptionalDefaults to 2000

Max output size in tokens, up to 4000

temperaturedoubleOptional

The temperature to use for the model. Higher values result in answers that are more creative, lower values are more conservative. Can be any value between 0.0 and 1.0 inclusive.

transcript_idslist of stringsOptional

A list of completed transcripts with text. Up to a maximum of 100 hours of audio. Use either transcript_ids or input_text as input into LeMUR.

Response

LeMUR summary response

request_idstringformat: "uuid"

The ID of the LeMUR request

responsestring

The response generated by LeMUR.

usageobject

The usage numbers for the LeMUR request

Authentication

Request

Response

Errors