Track costs

Understand automatic and custom cost tracking for LLM operations in Weave

Weave tracks the cost of LLM calls in two ways:

• Automatic cost tracking: For supported integrations, Weave captures token usage from the API response and applies built-in pricing for the model, with no extra code required.
• Custom cost tracking: For fine-tuned models, self-hosted models, or models that Weave doesn’t automatically integrate with, configure Weave to track custom costs using available API methods.

The Weave TypeScript does not support cost tracking at this time.

Use automatic cost tracking

When you call weave.init() and use a supported LLM integration such as OpenAI, Anthropic, Cohere, or Mistral, Weave automatically records token usage and calculates cost for each call. Costs appear in the trace tree and in the calls table in the Weave UI, and are available programmatically under call.summary["weave"]["costs"] when you query calls with the include_costs parameter set to true.

Automatic cost tracking requires two conditions:

• The LLM provider is a supported integration.
• The API response includes token usage (most providers return this by default).

If either condition is not met, use custom cost tracking instead.

The following example shows how to retrieve automatic cost data programmatically:

import weave
from openai import OpenAI

weave.init("your-team/project-name")
client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "What is 2 + 2?"}],
)

import time
time.sleep(2)

# Retrieve the current instance of the Weave client
weave_client = weave.get_client()
# Access the calls for the instance and their costs
calls = list(weave_client.get_calls(include_costs=True, limit=1))
call = calls[0]

# Access the call summary and retrieve the available costs fields
costs = call.summary.get("weave", {}).get("costs", {})
if not costs:
    print("No costs found in summary.weave.costs")
for model, cost in costs.items():
    print(f"Model: {model}")
    print(f"  Input cost:  ${cost['prompt_tokens_total_cost']:.6f}")
    print(f"  Output cost: ${cost['completion_tokens_total_cost']:.6f}")

Add a custom cost

You can add a custom cost by using the add_cost method. The three required fields are llm_id, prompt_token_cost, and completion_token_cost. llm_id is the name of the LLM (for example, gpt-4o). prompt_token_cost and completion_token_cost are cost per token for the LLM (if the LLM prices were specified in per million tokens, make sure to convert the value). You can also set effective_date to a datetime, to make the cost effective at a specific date, this defaults to the current date.

import weave
from datetime import datetime

client = weave.init("your-team/project-name")

client.add_cost(
    llm_id="your_model_name",
    prompt_token_cost=0.01,
    completion_token_cost=0.02
)

client.add_cost(
    llm_id="your_model_name",
    prompt_token_cost=10,
    completion_token_cost=20,
    effective_date=datetime(2025, 4, 22),
)

Querying custom costs

You can query for costs by using the query_costs method. There are a few ways to query for costs, you can pass in a singular cost id, or a list of LLM model names.

import weave

client = weave.init("your-team/project-name")

costs = client.query_costs(llm_ids=["your_model_name"])

cost = client.query_costs(costs[0].id)

Purging a custom cost

You can purge a custom cost by using the purge_costs method. You pass in a list of cost ids, and the costs with those ids are purged.

import weave

client = weave.init("your-team/project-name")

costs = client.query_costs(llm_ids=["your_model_name"])
client.purge_costs([cost.id for cost in costs])

Calculating custom costs for a project

You can calculate costs for a project by using get_calls() with include_costs=True.

import weave

weave.init("your-team/project-name")

@weave.op()
def get_costs_for_project(project_name: str):
    total_cost = 0
    requests = 0

    client = weave.init(project_name)
    calls = list(
        client.get_calls(filter={"trace_roots_only": True}, include_costs=True)
    )

    for call in calls:
        if call.summary["weave"] is not None and call.summary["weave"].get("costs", None) is not None:
            for k, cost in call.summary["weave"]["costs"].items():
                requests += cost["requests"]
                total_cost += cost["prompt_tokens_total_cost"]
                total_cost += cost["completion_tokens_total_cost"]

    return {
        "total_cost": total_cost,
        "requests": requests,
        "calls": len(calls),
    }

get_costs_for_project("my_custom_cost_model")

Setting up a custom model with custom costs

Try the cookbook for setting up costs with a custom model.

Try in Colab