Anthropic SDK Python v0.97.0 introduces CMA Memory public beta, fixes API specification errors, and optimizes multipart request performance. The update was released on for Python developers using Anthropic’s API services.
| Released by | Anthropic |
|---|---|
| Release date | |
| What it is | Python SDK update with CMA Memory beta feature |
| Who it’s for | Python developers using Anthropic’s API |
| Where to get it | GitHub repository |
| Price | Free |
- CMA Memory feature enters public beta for enhanced AI conversation context
- API specification errors fixed to improve developer experience
- Multipart request file structure copying optimized for better performance
- Missing features restored from previous SDK versions
- Bootstrap script made more robust for installation reliability
- CMA Memory beta enables persistent conversation context across API sessions
- Performance improvements reduce overhead in file upload operations
- API specification fixes enhance SDK reliability and developer productivity
- Update maintains backward compatibility while adding new capabilities
What is Anthropic SDK Python
Anthropic SDK Python is the official Python library for integrating with Anthropic’s Claude AI models and services. The SDK provides developers with streamlined access to Anthropic’s API endpoints, handling authentication, request formatting, and response parsing. It supports both synchronous and asynchronous operations for building AI-powered applications.
The SDK abstracts complex API interactions into simple Python methods, enabling developers to integrate Claude’s capabilities without managing low-level HTTP requests. It includes type hints, error handling, and automatic retry logic for production-ready applications.
What is new vs the previous version
Version 0.97.0 introduces CMA Memory public beta as the primary new feature compared to v0.96.0.
| Feature Category | v0.96.0 | v0.97.0 |
|---|---|---|
| CMA Memory | Not available | Public beta |
| API Specification | Contains errors | Fixed errors |
| Multipart Requests | Standard performance | Optimized file copying |
| Feature Completeness | Missing some features | Restored missing features |
| Bootstrap Script | Basic robustness | Enhanced robustness |
How does CMA Memory work
CMA Memory enables persistent conversation context that maintains information across multiple API interactions. Not yet disclosed how the memory system stores and retrieves contextual information.
- Initialize CMA Memory session through the SDK’s new beta endpoints
- Submit conversation context that gets stored in the memory system
- Reference stored memories in subsequent API calls for enhanced responses
- Memory persists across sessions until explicitly cleared or expired
Benchmarks and evidence
Performance improvements focus on multipart request optimization, though specific metrics are not yet disclosed.
| Metric | Improvement | Source |
|---|---|---|
| File structure copying speed | Optimized | GitHub commit 1f9eed3 |
| API specification accuracy | Errors fixed | GitHub commit f946de8 |
| Feature availability | Missing features restored | GitHub commit 72212ab |
Who should care
Builders
Python developers building AI applications benefit from CMA Memory’s persistent context capabilities. The performance optimizations reduce latency in file upload operations, improving user experience in document processing applications.
Enterprise
Enterprise teams using Anthropic’s API for production applications gain more reliable SDK behavior through bug fixes. CMA Memory beta enables more sophisticated conversational AI implementations for customer service and internal tools.
End users
End users experience improved application performance and more contextually aware AI interactions. Applications built with the updated SDK provide smoother file upload experiences and better conversation continuity.
Investors
The CMA Memory feature represents Anthropic’s advancement in persistent AI context management, potentially differentiating their offerings from competitors. Regular SDK updates demonstrate ongoing developer ecosystem investment.
How to use today
Install the updated SDK using pip package manager and access CMA Memory through beta endpoints.
- Update the SDK:
pip install anthropic==0.97.0 - Import the client:
from anthropic import Anthropic - Initialize with API key:
client = Anthropic(api_key="your-key") - Access CMA Memory beta features through the client’s new methods
- Test multipart file uploads to verify performance improvements
SDK vs competitors
Anthropic SDK Python competes with OpenAI’s Python SDK and other AI service libraries in developer adoption.
| Feature | Anthropic SDK v0.97.0 | OpenAI Python SDK | Google AI SDK |
|---|---|---|---|
| Memory persistence | CMA Memory beta | Not available | Not available |
| File upload optimization | Optimized in v0.97.0 | Standard | Standard |
| Type hints | Full support | Full support | Partial support |
| Async support | Yes | Yes | Yes |
Risks, limits, and myths
- CMA Memory is in public beta, so stability and feature completeness may vary
- Performance improvements apply specifically to multipart requests, not all operations
- API specification fixes may introduce breaking changes for edge case implementations
- Memory persistence duration and storage limits are not yet disclosed
- Beta features may change significantly before general availability
FAQ
How do I upgrade to Anthropic SDK Python v0.97.0?
Run pip install anthropic==0.97.0 to upgrade to the latest version with CMA Memory beta support.
What is CMA Memory in Anthropic SDK?
CMA Memory is a public beta feature that enables persistent conversation context across multiple API interactions for enhanced AI responses.
Does v0.97.0 break compatibility with previous versions?
The update maintains backward compatibility while adding new features and fixing API specification errors.
How much faster are multipart requests in v0.97.0?
Specific performance metrics for the multipart request optimization are not yet disclosed by Anthropic.
Can I use CMA Memory in production applications?
CMA Memory is in public beta, so evaluate stability requirements before using in production environments.
What API specification errors were fixed in v0.97.0?
Specific details about the API specification error fixes are not yet disclosed in the release notes.
How long does CMA Memory store conversation context?
Memory persistence duration and storage limits for CMA Memory are not yet disclosed.
Does the SDK support both sync and async operations?
Yes, Anthropic SDK Python supports both synchronous and asynchronous operations for different application architectures.
Glossary
- CMA Memory
- A beta feature enabling persistent conversation context storage across API sessions
- Multipart Request
- HTTP request format used for uploading files and complex data structures
- API Specification
- Technical documentation defining how to interact with an API’s endpoints and parameters
- SDK
- Software Development Kit providing tools and libraries for building applications
- Bootstrap Script
- Initialization script that sets up the development environment and dependencies
Sources
- Anthropic SDK Python v0.97.0 Release Notes. GitHub. . https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.97.0
