Rate limit exceed using Langfuse Python SDK

[laisalvescw]

Hello, I am facing an issue of rate limit exceed (error 429) when extracting traces using the following code:

response = langfuse.client.trace.list( name=name, page=page_num, user_id=user_id )

I am developing an external pipeline evaluation using Python, but I am not being able to eficientily extract Langfuse traces.
What is the limit for the API request? There is any other way to efficiently extract Langfuse Traces that meet a certain criteria, such as totalTokens > 1300, or total cost > 0.05?

[dosubot[bot]]

Hey @laisalvescw! 👋 I'm here to help you with any bugs, questions, or becoming a contributor. Let's tackle this together!

The rate limit for the Langfuse API when using the Python SDK is 1000 batches per minute for Hobby/Pro users and 5000 batches per minute for Team users ^[1].

To efficiently extract Langfuse traces and avoid hitting the rate limit, you can use the SDK's support for event queuing and batching. You can configure the batching behavior using the flush_at and flush_interval options. For example, set flushAt=1 to send every event immediately or flushInterval=1000 to send every second. You can also manually flush events using the flush method ^[2][3].

Regarding filtering traces by criteria such as totalTokens or total cost, the documentation does not specify built-in functionality for this in the Python SDK. However, you can fetch traces and apply filtering logic in your code after retrieving the data.

To continue talking to Dosu, mention @dosu.

[jannikmaierhoefer]

Hi @laisalvescw, dosu-bot is right. You can find our API rate limits for each Langfuse plan here: https://langfuse.com/faq/all/api-limits