langchain-ai/langchain

Add Perplexity cost tracking

Open

#31,647 opened on Apr 14, 2025

View on GitHub
 (6 comments) (1 reaction) (1 assignee)Python (136,758 stars) (22,617 forks)batch import
externalhelp wantedintegrationinvestigateperplexity

Description

Issue you'd like to raise.

costs shown on langsmith for perplexity API calls are inaccurate (even upto an order of magnitude). refer to their docs for how to calculate costs: https://docs.perplexity.ai/guides/models/sonar-reasoning-pro#detailed-pricing-breakdown-for-sonar-reasoning-pro

The problem is that pplx cost depends not only on input and output tokens but also on search queries and

Suggestion:

either add support for calculating this within langsmith platform or let users update the true cost while tracing an LLM call.

Contributor guide

Tech stack
python
Domain
backendanalytics
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
blocked
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
Basic PythonLangChain framework basicsAPI cost model understanding
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
35
Research direction
Review the Perplexity API pricing documentation to understand the cost formula (input tokens, output tokens, search queries). Examine the LangChain callback system (e.g., `libs/langchain/langchain/callbacks/tracers/`) for existing cost tracking logic. Check issue comments for maintainer input or assignee progress. Consider extending the LangSmith tracing client or adding a per call cost override.