Python SDK

The custom music compose_detailed wrapper now forwards the enable_logging flag.

What's Changed

• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/901

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.73.1...v2.74.0

What's Changed

• Fix tests on Python 3.13+, add 3.13/3.14 to CI, bump to 2.73.1 by @PaulAsjes in https://github.com/elevenlabs/elevenlabs-python/pull/898

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.73.0...v2.73.1

New create_and_wait helpers are available for Flows generations and template runs. Custom sub-client properties are now typed so pyright sees construct_event, realtime, and convert_realtime.

RealtimeEvents.CLOSE now includes the close code and reason, and play() and save() accept any iterable of audio chunks rather than requiring a specific type.

What's Changed

• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/888

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.70.0...v2.71.0

What's Changed

• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/881

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.69.0...v2.70.0

The realtime speech-to-text wrapper now exposes transcript_edit. Fixed ffplay exiting with an error silently rather than raising, and API keys with a residency suffix being stripped before JWT verification.

compose_detailed now accepts every music model id and composition plan type, replacing the previous narrower set of accepted values. The rest of the release is an SDK regeneration.

SDK v3 regeneration and hand-written migration cover agents rename, orchestrator config, and generated websocket clients. Also adds file_ids to send_multimodal_message.

Fixed an issue where sending the OMIT keyword during realtime TTS could disconnect the connection.

The send_multimodal_message method now accepts file_ids parameter.

What's Changed

• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/852

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.64.0...v2.65.0

Adds an extra_setup_config escape hatch to customize the on-prem setup message.

Added the missing realtime speech-to-text options to the SDK.

What's Changed

• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/835

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.61.0...v2.62.0

What's Changed

• chore: forget stale Fern Replay patch for #806 (recurring version-bump conflict) by @kraenhansen in https://github.com/elevenlabs/elevenlabs-python/pull/828
• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/833

Fixed a blocking issue in AsyncConversation startup and corrected how list-of-primitive multipart fields are sent as repeated form fields.

Added HMAC-signed post-call webhook configuration for on-premises conversations.

What's Changed

• SDK regeneration by @fern-api[bot] in https://github.com/elevenlabs/elevenlabs-python/pull/810

Full Changelog: https://github.com/elevenlabs/elevenlabs-python/compare/v2.57.0...v2.58.0