# AgentMeter Python SDK v0.3.1
**Modern Python SDK for usage tracking and billing in AI applications.**
[](https://pypi.org/project/agentmeter/)
[](https://pypi.org/project/agentmeter/)
[](https://opensource.org/licenses/MIT)
## ✨ Features
- 🎯 **Resource-Based Architecture** - Clean, organized API with dedicated resource managers
- ⚡ **Async/Await Support** - Full async support for modern Python applications
- 🔌 **Rich Integrations** - First-class support for LangChain, Coinbase AgentKit, and more
- 🛡️ **Type Safety** - Built with Pydantic v2 for robust data validation and type hints
- 🌐 **Production Ready** - Enterprise-grade reliability with retry logic and error handling
- 📊 **Flexible Billing** - Support for usage-based, token-based, and instant billing models
- 🔄 **100% Backward Compatible** - All v0.2.0 code continues to work unchanged
## 🚀 Quick Start
### Installation
```bash
pip install agentmeter
```
### Basic Usage
```python
from agentmeter import AgentMeter
# Initialize client
with AgentMeter(api_key="your_api_key") as meter:
# Create a meter type
meter_type = meter.meter_types.create(
name="api_calls",
unit="requests",
description="API usage tracking"
)
# Record usage event
event = meter.meter_events.record(
meter_type_id=meter_type.id,
subject_id="user_123",
quantity=1.0,
metadata={"endpoint": "/api/search"}
)
# Get usage statistics
stats = meter.meter_events.get_aggregations(
meter_type_id=meter_type.id,
subject_id="user_123"
)
print(f"Total usage: {stats.total_quantity} {meter_type.unit}")
```
## 🏗️ Core Concepts
### Resource Managers
AgentMeter v0.3.1 organizes functionality into focused resource managers:
```python
meter.projects # Project management
meter.meter_types # Meter type definitions
meter.meter_events # Usage event recording and querying
meter.users # User-specific meter management
```
### Async Support
```python
from agentmeter import AsyncAgentMeter
async with AsyncAgentMeter(api_key="your_api_key") as meter:
# All methods have async equivalents
meter_type = await meter.meter_types.create(...)
event = await meter.meter_events.record(...)
stats = await meter.meter_events.get_aggregations(...)
```
## 🔌 Integrations
### 🚀 Coinbase AgentKit
Build monetized Web3 AI agents:
```python
from agentmeter import AgentMeter
meter = AgentMeter(api_key="your_api_key")
# Create meter types for different Web3 operations
wallet_queries = meter.meter_types.create(
name="wallet_queries",
unit="queries",
description="Blockchain wallet queries"
)
trading_operations = meter.meter_types.create(
name="trading_ops",
unit="trades",
description="Smart contract trading"
)
# Track usage in your AgentKit application
def get_wallet_balance(user_id: str, asset: str):
# Your Web3 logic here
balance = wallet.get_balance(asset)
# Record billable event
meter.meter_events.record(
meter_type_id=wallet_queries.id,
subject_id=user_id,
quantity=1.0,
metadata={"asset": asset, "operation": "balance_check"}
)
return balance
```
**[📖 Full Coinbase AgentKit Integration Example](examples/coinbase_agentkit_integration.py)**
### 🤖 LangChain
Automatic LLM usage tracking:
```python
from agentmeter.langchain_integration import LangChainAgentMeterCallback
# Create callback with your meter
callback = LangChainAgentMeterCallback(
agentmeter_client=meter,
meter_type_id="llm_usage_meter_id"
)
# Add to any LangChain component
llm = ChatOpenAI(callbacks=[callback])
chain = ConversationChain(llm=llm, callbacks=[callback])
# Usage automatically tracked
result = chain.run("Analyze this data...")
```
**[📖 Full LangChain Integration Example](examples/langchain_integration_meter.py)**
## 💰 Billing Models
### Usage-Based Billing
Perfect for API calls, processing requests, or resource consumption:
```python
# Create usage meter
api_meter = meter.meter_types.create(
name="api_requests",
unit="requests"
)
# Record usage
meter.meter_events.record(
meter_type_id=api_meter.id,
subject_id="user_123",
quantity=1.0
)
```
### Token-Based Billing
Ideal for AI/LLM applications:
```python
# Create token meter
token_meter = meter.meter_types.create(
name="llm_tokens",
unit="tokens"
)
# Record token usage
meter.meter_events.record(
meter_type_id=token_meter.id,
subject_id="user_123",
quantity=1500.0, # Total tokens used
metadata={
"input_tokens": 1000,
"output_tokens": 500,
"model": "gpt-4"
}
)
```
### Event-Based Billing
For feature unlocks or one-time charges:
```python
# Create event meter
feature_meter = meter.meter_types.create(
name="premium_features",
unit="unlocks"
)
# Record feature usage
meter.meter_events.record(
meter_type_id=feature_meter.id,
subject_id="user_123",
quantity=1.0,
metadata={
"feature": "advanced_analytics",
"tier": "premium"
}
)
```
## 📊 Analytics & Reporting
### Usage Aggregations
```python
# Get aggregated usage statistics
stats = meter.meter_events.get_aggregations(
meter_type_id=meter_type.id,
subject_id="user_123",
start_time="2024-01-01T00:00:00Z",
end_time="2024-01-31T23:59:59Z"
)
print(f"Total usage: {stats.total_quantity}")
print(f"Event count: {stats.total_events}")
print(f"Period: {stats.period_start} to {stats.period_end}")
```
### User Meter Management
```python
# Set user spending limits
user_meter = meter.users.set_meter(
user_id="user_123",
threshold_amount=100.0 # $100 monthly limit
)
# Check current usage
current_usage = meter.users.get_meter(user_id="user_123")
print(f"Usage: ${current_usage.current_usage} / ${current_usage.threshold_amount}")
# Reset usage (e.g., monthly reset)
meter.users.reset_meter(user_id="user_123")
```
## 🛠️ Configuration
### Environment Variables
```bash
export AGENTMETER_API_KEY="your_api_key"
export AGENTMETER_BASE_URL="https://api.agentmeter.money" # Optional
```
### Programmatic Configuration
```python
from agentmeter import AgentMeter
meter = AgentMeter(
api_key="your_api_key",
base_url="https://api.agentmeter.money", # Optional
timeout=30.0,
max_retries=3
)
```
## 🔄 Migration from v0.2.0
AgentMeter v0.3.1 maintains 100% backward compatibility. Your existing code continues to work:
```python
# ✅ This still works (legacy API)
from agentmeter import AgentMeterClient
client = AgentMeterClient(api_key="key")
client.record_api_request_pay(api_calls=1, unit_price=0.10)
# ✨ New recommended approach (v0.3.1)
from agentmeter import AgentMeter
with AgentMeter(api_key="key") as meter:
meter.meter_events.record(meter_type_id="mt_123", subject_id="user", quantity=1.0)
```
For a smooth migration:
1. **Keep existing code working** - No immediate changes required
2. **Gradually adopt new API** - Use v0.3.1 for new features
3. **Migrate when convenient** - Update modules one at a time
**[📖 See complete examples: v0.3.1 vs Legacy](examples/v031_new_features.py)**
## 🔧 Error Handling
```python
from agentmeter import AgentMeter
from agentmeter.exceptions import AuthenticationError, ValidationError, ServerError
try:
with AgentMeter(api_key="invalid") as meter:
event = meter.meter_events.record(...)
except AuthenticationError:
print("Invalid API key")
except ValidationError as e:
print(f"Data validation failed: {e}")
except ServerError as e:
print(f"Server error: {e}")
```
## 🧪 Examples & Testing
```bash
# Run the new v0.3.1 examples
python examples/v031_new_features.py
python examples/coinbase_agentkit_integration.py
python examples/langchain_integration_meter.py
python examples/ecommerce_integration.py
# Run tests
pytest tests/
```
## 📚 Documentation
- **[API Reference](https://docs.agentmeter.money/api)**
- **[Integration Guides](https://docs.agentmeter.money/integrations)**
- **[🗂️ Legacy Documentation (v0.2.0 and earlier)](HISTORICAL_VERSIONS.md)**
## 🤝 Support
- **[GitHub Issues](https://github.com/Pagent-Money/agentmeter-sdk-python/issues)**
- **[Documentation](https://docs.agentmeter.money)**
- **[Email Support](mailto:thomas.yu@knn3.xyz)**
## 📄 License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
---
**🚀 Ready to get started? Install AgentMeter and start tracking usage in minutes!**
```bash
pip install agentmeter
```
Raw data
{
"_id": null,
"home_page": null,
"name": "agentmeter",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": null,
"keywords": "billing, usage-tracking, metering, api-billing, token-billing, subscription, payments, agentmeter",
"author": null,
"author_email": "\"AgentMeter.Money\" <thomas.yu@knn3.xyz>",
"download_url": "https://files.pythonhosted.org/packages/10/3f/4cf4581aef74720e7bc13782046dcd360e476fb308a4c16fe7d31d65da3e/agentmeter-0.3.2.tar.gz",
"platform": null,
"description": "\n\n# AgentMeter Python SDK v0.3.1\n\n**Modern Python SDK for usage tracking and billing in AI applications.**\n\n[](https://pypi.org/project/agentmeter/)\n[](https://pypi.org/project/agentmeter/)\n[](https://opensource.org/licenses/MIT)\n\n## \u2728 Features\n\n- \ud83c\udfaf **Resource-Based Architecture** - Clean, organized API with dedicated resource managers\n- \u26a1 **Async/Await Support** - Full async support for modern Python applications\n- \ud83d\udd0c **Rich Integrations** - First-class support for LangChain, Coinbase AgentKit, and more\n- \ud83d\udee1\ufe0f **Type Safety** - Built with Pydantic v2 for robust data validation and type hints\n- \ud83c\udf10 **Production Ready** - Enterprise-grade reliability with retry logic and error handling\n- \ud83d\udcca **Flexible Billing** - Support for usage-based, token-based, and instant billing models\n- \ud83d\udd04 **100% Backward Compatible** - All v0.2.0 code continues to work unchanged\n\n## \ud83d\ude80 Quick Start\n\n### Installation\n\n```bash\npip install agentmeter\n```\n\n### Basic Usage\n\n```python\nfrom agentmeter import AgentMeter\n\n# Initialize client\nwith AgentMeter(api_key=\"your_api_key\") as meter:\n # Create a meter type\n meter_type = meter.meter_types.create(\n name=\"api_calls\",\n unit=\"requests\",\n description=\"API usage tracking\"\n )\n \n # Record usage event\n event = meter.meter_events.record(\n meter_type_id=meter_type.id,\n subject_id=\"user_123\",\n quantity=1.0,\n metadata={\"endpoint\": \"/api/search\"}\n )\n \n # Get usage statistics\n stats = meter.meter_events.get_aggregations(\n meter_type_id=meter_type.id,\n subject_id=\"user_123\"\n )\n print(f\"Total usage: {stats.total_quantity} {meter_type.unit}\")\n```\n\n## \ud83c\udfd7\ufe0f Core Concepts\n\n### Resource Managers\n\nAgentMeter v0.3.1 organizes functionality into focused resource managers:\n\n```python\nmeter.projects # Project management\nmeter.meter_types # Meter type definitions \nmeter.meter_events # Usage event recording and querying\nmeter.users # User-specific meter management\n```\n\n### Async Support\n\n```python\nfrom agentmeter import AsyncAgentMeter\n\nasync with AsyncAgentMeter(api_key=\"your_api_key\") as meter:\n # All methods have async equivalents\n meter_type = await meter.meter_types.create(...)\n event = await meter.meter_events.record(...)\n stats = await meter.meter_events.get_aggregations(...)\n```\n\n## \ud83d\udd0c Integrations\n\n### \ud83d\ude80 Coinbase AgentKit\n\nBuild monetized Web3 AI agents:\n\n```python\nfrom agentmeter import AgentMeter\n\nmeter = AgentMeter(api_key=\"your_api_key\") \n\n# Create meter types for different Web3 operations\nwallet_queries = meter.meter_types.create(\n name=\"wallet_queries\", \n unit=\"queries\",\n description=\"Blockchain wallet queries\"\n)\n\ntrading_operations = meter.meter_types.create(\n name=\"trading_ops\",\n unit=\"trades\", \n description=\"Smart contract trading\"\n)\n\n# Track usage in your AgentKit application\ndef get_wallet_balance(user_id: str, asset: str):\n # Your Web3 logic here\n balance = wallet.get_balance(asset)\n \n # Record billable event\n meter.meter_events.record(\n meter_type_id=wallet_queries.id,\n subject_id=user_id,\n quantity=1.0,\n metadata={\"asset\": asset, \"operation\": \"balance_check\"}\n )\n \n return balance\n```\n\n**[\ud83d\udcd6 Full Coinbase AgentKit Integration Example](examples/coinbase_agentkit_integration.py)**\n\n### \ud83e\udd16 LangChain\n\nAutomatic LLM usage tracking:\n\n```python\nfrom agentmeter.langchain_integration import LangChainAgentMeterCallback\n\n# Create callback with your meter\ncallback = LangChainAgentMeterCallback(\n agentmeter_client=meter,\n meter_type_id=\"llm_usage_meter_id\"\n)\n\n# Add to any LangChain component\nllm = ChatOpenAI(callbacks=[callback])\nchain = ConversationChain(llm=llm, callbacks=[callback])\n\n# Usage automatically tracked\nresult = chain.run(\"Analyze this data...\")\n```\n\n**[\ud83d\udcd6 Full LangChain Integration Example](examples/langchain_integration_meter.py)**\n\n## \ud83d\udcb0 Billing Models\n\n### Usage-Based Billing\n\nPerfect for API calls, processing requests, or resource consumption:\n\n```python\n# Create usage meter\napi_meter = meter.meter_types.create(\n name=\"api_requests\",\n unit=\"requests\"\n)\n\n# Record usage\nmeter.meter_events.record(\n meter_type_id=api_meter.id,\n subject_id=\"user_123\", \n quantity=1.0\n)\n```\n\n### Token-Based Billing\n\nIdeal for AI/LLM applications:\n\n```python\n# Create token meter\ntoken_meter = meter.meter_types.create(\n name=\"llm_tokens\",\n unit=\"tokens\"\n)\n\n# Record token usage\nmeter.meter_events.record(\n meter_type_id=token_meter.id,\n subject_id=\"user_123\",\n quantity=1500.0, # Total tokens used\n metadata={\n \"input_tokens\": 1000,\n \"output_tokens\": 500,\n \"model\": \"gpt-4\"\n }\n)\n```\n\n### Event-Based Billing\n\nFor feature unlocks or one-time charges:\n\n```python\n# Create event meter \nfeature_meter = meter.meter_types.create(\n name=\"premium_features\",\n unit=\"unlocks\"\n)\n\n# Record feature usage\nmeter.meter_events.record(\n meter_type_id=feature_meter.id,\n subject_id=\"user_123\",\n quantity=1.0,\n metadata={\n \"feature\": \"advanced_analytics\",\n \"tier\": \"premium\"\n }\n)\n```\n\n## \ud83d\udcca Analytics & Reporting\n\n### Usage Aggregations\n\n```python\n# Get aggregated usage statistics\nstats = meter.meter_events.get_aggregations(\n meter_type_id=meter_type.id,\n subject_id=\"user_123\",\n start_time=\"2024-01-01T00:00:00Z\",\n end_time=\"2024-01-31T23:59:59Z\"\n)\n\nprint(f\"Total usage: {stats.total_quantity}\")\nprint(f\"Event count: {stats.total_events}\")\nprint(f\"Period: {stats.period_start} to {stats.period_end}\")\n```\n\n### User Meter Management\n\n```python\n# Set user spending limits\nuser_meter = meter.users.set_meter(\n user_id=\"user_123\",\n threshold_amount=100.0 # $100 monthly limit\n)\n\n# Check current usage\ncurrent_usage = meter.users.get_meter(user_id=\"user_123\")\nprint(f\"Usage: ${current_usage.current_usage} / ${current_usage.threshold_amount}\")\n\n# Reset usage (e.g., monthly reset)\nmeter.users.reset_meter(user_id=\"user_123\")\n```\n\n## \ud83d\udee0\ufe0f Configuration\n\n### Environment Variables\n\n```bash\nexport AGENTMETER_API_KEY=\"your_api_key\"\nexport AGENTMETER_BASE_URL=\"https://api.agentmeter.money\" # Optional\n```\n\n### Programmatic Configuration\n\n```python\nfrom agentmeter import AgentMeter\n\nmeter = AgentMeter(\n api_key=\"your_api_key\",\n base_url=\"https://api.agentmeter.money\", # Optional\n timeout=30.0,\n max_retries=3\n)\n```\n\n## \ud83d\udd04 Migration from v0.2.0\n\nAgentMeter v0.3.1 maintains 100% backward compatibility. Your existing code continues to work:\n\n```python\n# \u2705 This still works (legacy API)\nfrom agentmeter import AgentMeterClient\nclient = AgentMeterClient(api_key=\"key\")\nclient.record_api_request_pay(api_calls=1, unit_price=0.10)\n\n# \u2728 New recommended approach (v0.3.1)\nfrom agentmeter import AgentMeter\nwith AgentMeter(api_key=\"key\") as meter:\n meter.meter_events.record(meter_type_id=\"mt_123\", subject_id=\"user\", quantity=1.0)\n```\n\nFor a smooth migration:\n\n1. **Keep existing code working** - No immediate changes required\n2. **Gradually adopt new API** - Use v0.3.1 for new features\n3. **Migrate when convenient** - Update modules one at a time\n\n**[\ud83d\udcd6 See complete examples: v0.3.1 vs Legacy](examples/v031_new_features.py)**\n\n## \ud83d\udd27 Error Handling\n\n```python\nfrom agentmeter import AgentMeter\nfrom agentmeter.exceptions import AuthenticationError, ValidationError, ServerError\n\ntry:\n with AgentMeter(api_key=\"invalid\") as meter:\n event = meter.meter_events.record(...)\n \nexcept AuthenticationError:\n print(\"Invalid API key\")\nexcept ValidationError as e:\n print(f\"Data validation failed: {e}\")\nexcept ServerError as e:\n print(f\"Server error: {e}\")\n```\n\n## \ud83e\uddea Examples & Testing\n\n```bash\n# Run the new v0.3.1 examples\npython examples/v031_new_features.py\npython examples/coinbase_agentkit_integration.py \npython examples/langchain_integration_meter.py\npython examples/ecommerce_integration.py\n\n# Run tests\npytest tests/\n```\n\n## \ud83d\udcda Documentation\n\n- **[API Reference](https://docs.agentmeter.money/api)**\n- **[Integration Guides](https://docs.agentmeter.money/integrations)**\n- **[\ud83d\uddc2\ufe0f Legacy Documentation (v0.2.0 and earlier)](HISTORICAL_VERSIONS.md)**\n\n## \ud83e\udd1d Support\n\n- **[GitHub Issues](https://github.com/Pagent-Money/agentmeter-sdk-python/issues)**\n- **[Documentation](https://docs.agentmeter.money)**\n- **[Email Support](mailto:thomas.yu@knn3.xyz)**\n\n## \ud83d\udcc4 License\n\nThis project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.\n\n---\n\n**\ud83d\ude80 Ready to get started? Install AgentMeter and start tracking usage in minutes!**\n\n```bash\npip install agentmeter\n```\n",
"bugtrack_url": null,
"license": null,
"summary": "Python SDK for AgentMeter usage tracking and billing",
"version": "0.3.2",
"project_urls": {
"Changelog": "https://github.com/Pagent-Money/agentmeter-sdk-python/blob/main/CHANGELOG.md",
"Documentation": "https://docs.agentmeter.money/sdk/python",
"Homepage": "https://agentmeter.money",
"Issues": "https://github.com/Pagent-Money/agentmeter-sdk-python/issues",
"Repository": "https://github.com/Pagent-Money/agentmeter-sdk-python"
},
"split_keywords": [
"billing",
" usage-tracking",
" metering",
" api-billing",
" token-billing",
" subscription",
" payments",
" agentmeter"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "8f8fceb180191c385c6396087baa5b7b36210aef49583f1cbfef57a491b14208",
"md5": "28d445c44f5ab702107955b6903bbb4d",
"sha256": "e54143074bb2318fdc507f0d6fbb7607122510bf4907f82186d08c4c294ee447"
},
"downloads": -1,
"filename": "agentmeter-0.3.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "28d445c44f5ab702107955b6903bbb4d",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 29799,
"upload_time": "2025-07-10T08:16:44",
"upload_time_iso_8601": "2025-07-10T08:16:44.242815Z",
"url": "https://files.pythonhosted.org/packages/8f/8f/ceb180191c385c6396087baa5b7b36210aef49583f1cbfef57a491b14208/agentmeter-0.3.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "103f4cf4581aef74720e7bc13782046dcd360e476fb308a4c16fe7d31d65da3e",
"md5": "621b7027c37ed9d6307627592d18e476",
"sha256": "2586f0c453abd21deb6d4deb4099a04e47fe1eca592d5d23a5889de6cc91bb9e"
},
"downloads": -1,
"filename": "agentmeter-0.3.2.tar.gz",
"has_sig": false,
"md5_digest": "621b7027c37ed9d6307627592d18e476",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 66369,
"upload_time": "2025-07-10T08:16:45",
"upload_time_iso_8601": "2025-07-10T08:16:45.867583Z",
"url": "https://files.pythonhosted.org/packages/10/3f/4cf4581aef74720e7bc13782046dcd360e476fb308a4c16fe7d31d65da3e/agentmeter-0.3.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-07-10 08:16:45",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "Pagent-Money",
"github_project": "agentmeter-sdk-python",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [
{
"name": "httpx",
"specs": [
[
">=",
"0.24.0"
]
]
},
{
"name": "langchain",
"specs": [
[
">=",
"0.1.0"
]
]
},
{
"name": "langchain-core",
"specs": [
[
">=",
"0.1.0"
]
]
},
{
"name": "pydantic",
"specs": [
[
">=",
"2.0.0"
]
]
},
{
"name": "typing-extensions",
"specs": [
[
">=",
"4.5.0"
]
]
}
],
"lcname": "agentmeter"
}
JSON
