Anthropic Python SDK v1.15.0 Adds Speed Control and New Capabilities
The latest update to the anthropic-sdk-python introduces a speed parameter to messages, expands model capability metadata, and improves organizational spend tracking.
What changed?
Anthropic released v1.15.0 of its Python SDK. Major API updates include the ability to set a speed parameter on message requests and track usage speed, support for a new 'between_tools' thinking mode in model capabilities, and the exposure of an organization's usage-tier spend cap as a read-only limit. Some models require a beta header to use the speed parameter.
Why does it matter to an everyday developer?
The speed parameter enables developers to adjust the model's response speed, offering a way to balance cost, latency, or quality for different use cases. The addition of 'between_tools' in model capabilities helps in building agents that need to chain or reason between tool calls. The new spend cap reporting gives organizations more visibility into API spend and can help prevent unexpected charges.
What can the developer do now?
- Use the speed parameter in message requests to control generation latency, where supported.
- Check model capabilities for 'between_tools' to design more advanced tool-using agents.
- Monitor your organization’s spend cap via API for improved budget management.
- If targeting models that require the speed parameter, use the necessary beta header as described in the SDK documentation.