ElevenAgents: Conversation guardrail filters: List conversations adds optional guardrail_types…
By Steven Van ·
Conversation guardrail filters: List conversations adds optional guardrail_types (GuardrailType[]) and custom_guardrail_names (string array) query…
ElevenLabs has updated the ElevenLabs ElevenAgents API with several additions for developers building conversational agents, detailed in the August 10 changelog.
- The list conversations endpoint can now filter by guardrail_types, custom_guardrail_names, and parent_conversation_id.
- List agent branches adds an optional include_commit_status flag, which returns nullable commits_ahead, commits_behind, and merged_into_branch_id fields for each branch.
- Client overrides now support max_duration_seconds and pronunciation dictionary locators, MCP server configuration adds request_meta for passing MCP values to tool calls, and webhook tool timeouts can now go up to 300 seconds.
- Conversation charging now includes a per-feature analysis breakdown with running totals and run snapshots, transcript messages add a triggered_guardrails field, and conversation turns can include an optional producing_llm field.
ElevenLabs
AI audio platform for text to speech, voice cloning, and video dubbing across dozens of languages while preserving the original tone and emotion.
View ElevenLabs →