ElevenAgents Adds Parallel Tool Calls, Slack Alerting, and a GPT-6-Astra Option

ElevenLabs updated its ElevenAgents platform on September 21, 2026, with three changes aimed at making voice agents faster and easier to monitor in production: parallel tool execution, a new alerting channel, and an additional LLM option.

Details

• Parallel tool calls: agents built on supported models can now execute multiple tools within a single conversational turn, controlled by a new enable_parallel_tool_calls boolean (default true), instead of calling tools one at a time
• gpt-6-astra added: the model is now selectable as an agent LLM alongside ElevenAgents’ existing options
• Slack alerting: agent alerting now supports Slack channels as a notification target, joining the existing PagerDuty and generic webhook notifiers
• Phone number search: a new cursor-paginated endpoint lets teams filter phone numbers by provider, outbound support, assigned agent, branch, and labels

What happened next

The changes are incremental additions to ElevenAgents’ existing tool-calling and operations tooling rather than a standalone launch, but they matter for teams running agents at scale: parallel tool calls should cut latency on turns that need more than one lookup or action, and Slack alerting brings agent monitoring into a channel many support and ops teams already watch alongside PagerDuty. The updates shipped in the same week ElevenLabs also revised its Speech Engine retry timeout and turn-boundary markers, part of a steady cadence of infrastructure-level changelog updates the company has kept up through September.