Skip to content

Usage

LLMfy has a built-in usage tracker. Usage tracker is used to track token usage, request and cost. You can use llmfy_usage_tracker that imported from:

from llmfy import llmfy_usage_tracker

Use as context manager

llmfy_usage_tracker can be used to all supported providers on llmfy.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
with llmfy_usage_tracker() as usage:
    # Use chat or invoke
    # (chat use messages)
    response_a = agent.chat(messages, info=info)
    # (invoke use contents)
    response_b = agent.invoke(contents, info=info)

print(f"\n>> {response_a.result.content}\n")
print(f"\n>> {response_b.result.content}\n")
print(f"Usage:\n{usage}\n")

The output should be like this:

>> Maaf.


>> Maaf.

Usage:

------------------
USAGE: 

Total Tokens: 173
        Input Tokens: 167
        Output Tokens: 6
Total Requests: 2
Total Cost (USD): $2.865e-05
Total Cost (USD formatted): $0.00002865

Request Details:
1. gpt-4o-mini 
        provider: OPENAI 
        input_tokens: 60 
        output_tokens: 3 
        total_tokens: 63 
        input_price: 0.15 
        output_price: 0.6 
        price_per_tokens: 1000000 
        total_cost (USD): 1.08e-05 
        total_cost (USD formatted): 0.0000108

2. gpt-4o-mini 
        provider: OPENAI 
        input_tokens: 107 
        output_tokens: 3 
        total_tokens: 110 
        input_price: 0.15 
        output_price: 0.6 
        price_per_tokens: 1000000 
        total_cost (USD): 1.785e-05 
        total_cost (USD formatted): 0.00001785

------------------

Customize Prices Data

If the model is not found in the lmfy usage tracker, you can customize the model price source list based on the provider.

OpenAI

OpenAI Pricing dictionary source.

Example pricing structure: 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
{
    "gpt-4o": {
        "input": 2.50,
        "output": 10.00
    },
    "gpt-4o-mini": {
        "input": 0.15,
        "output": 0.60
    },
    "gpt-3.5-turbo": {
        "input": 0.05,
        "output": 1.50
    }
}

Bedrock

Bedrock Pricing dictionary source.

Example pricing structure: 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
{
    "anthropic.claude-3-5-sonnet-20240620-v1:0": {
        "us-east-1": {
            "region": "US East (N. Virginia)",
            "input": 0.003,
            "output": 0.015,
        },
        "us-west-2": {
            "region": "US West (Oregon)",
            "input": 0.003,
            "output": 0.015,
        },
    },
    "anthropic.claude-3-5-sonnet-20241022-v2:0": {
        "us-east-1": {
            "region": "US East (N. Virginia)",
            "input": 0.003,
            "output": 0.015,
        },
        "us-west-2": {
            "region": "US West (Oregon)",
            "input": 0.003,
            "output": 0.015,
        },
    }
}

Google AI

Google AI supports four pricing structures (all prices are per 1M tokens in USD):

Flat pricing — same price for all input types: 

1
2
3
4
5
6
{
    "gemini-2.0-flash": {
        "input": 0.10,
        "output": 0.40,
    }
}

Per-type input pricing — different price per modality: 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
{
    "gemini-3-flash-preview": {
        "input": {
            "default": 0.50,
            "text": 0.50,
            "image": 0.50,
            "video": 0.50,
            "audio": 1.00,
        },
        "output": 3.00,
    }
}

Tiered pricing — different price above a token count threshold: 

1
2
3
4
5
6
7
8
9
{
    "gemini-2.5-pro": {
        "input": 1.25,        # price when prompt <= threshold
        "input_high": 2.50,   # price when prompt > threshold
        "output": 10.00,
        "output_high": 15.00,
        "threshold": 200000,
    }
}

Tiered + per-type pricing — combines both tiered and per-modality pricing: 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
{
    "model-id": {
        "input": {
            "default": 0.25,
            "text": 0.25,
            "image": 0.25,
            "video": 0.25,
            "audio": 0.50,
        },
        "input_high": {
            "default": 0.50,
            "text": 0.50,
            "image": 0.50,
            "video": 0.50,
            "audio": 1.00,
        },
        "output": 1.50,
        "output_high": 3.00,
        "threshold": 200000,
    }
}

Example

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
openai_prices = {
    "gpt-4o": {
        "input": 2.50,
        "output": 10.00
    },
    "gpt-4o-mini": {
        "input": 0.15,
        "output": 0.60
    },
    "gpt-3.5-turbo": {
        "input": 0.05,
        "output": 1.50
    }
}

bedrock_prices = {
    "anthropic.claude-3-5-sonnet-20240620-v1:0": {
        "us-east-1": {
            "region": "US East (N. Virginia)",
            "input": 0.003,
            "output": 0.015,
        },
        "us-west-2": {
            "region": "US West (Oregon)",
            "input": 0.003,
            "output": 0.015,
        },
    },
    "anthropic.claude-3-5-sonnet-20241022-v2:0": {
        "us-east-1": {
            "region": "US East (N. Virginia)",
            "input": 0.003,
            "output": 0.015,
        },
        "us-west-2": {
            "region": "US West (Oregon)",
            "input": 0.003,
            "output": 0.015,
        },
    }
}

googleai_prices = {
    "gemini-2.0-flash": {
        "input": 0.10,
        "output": 0.40,
    },
    "gemini-2.5-pro": {
        "input": 1.25,
        "input_high": 2.50,
        "output": 10.00,
        "output_high": 15.00,
        "threshold": 200000,
    }
}

with llmfy_usage_tracker(
    openai_pricing=openai_prices,
    bedrock_pricing=bedrock_prices,
    googleai_pricing=googleai_prices,
) as usage:
    # invoke llmfy