The idea is:
Add an “Enable parallel tool use” option to the Anthropic Chat Models for agent node in n8n. This would expose Anthropic’s disable_parallel_tool_use flag (set to false when checked) so that multiple tools invoked in a single Sonnet prompt can be executed concurrently—just like how Google Gemini Flash 2.5 already handles tool calling in parallel within n8n.
My use case:
I build advanced workflows where Sonnet needs to invoke multiple tools—such as calling APIs, querying databases, or performing different logic tasks—all within a single prompt. Currently, these tool calls are executed one at a time, which adds unnecessary delay. With parallel tool use, these could run simultaneously, significantly improving execution speed and responsiveness.
I think it would be beneficial to add this because:
- Lower latency: Concurrent tool calls reduce total runtime, especially in multi-tool scenarios.
- Higher throughput: Workflows that process large volumes or complex branching logic will benefit from faster tool resolution.
- Parity with Gemini Flash 2.5: n8n’s Google Gemini node already supports parallel tool invocation—bringing the same to Sonnet ensures consistency across providers.
- Makes full use of Sonnet’s capabilities: Anthropic has built native support for parallel tool use; this flag simply unlocks it in n8n.
Any resources to support this?
- Anthropic’s official documentation on parallel tool use:
https://docs.anthropic.com/en/docs/build-with-claude/tool-use/implement-tool-use#parallel-tool-use
Are you willing to work on this?
Yes — I’m happy to test early builds, validate parallel performance, and provide feedback during rollout.