> This is a page from the ElevenLabs documentation. For a complete page index, fetch https://elevenlabs.io/docs/llms.txt. For the full documentation in a single file, fetch https://elevenlabs.io/docs/llms-full.txt.

import { YoutubeEmbed } from "../../../components/YoutubeEmbed";

### Eleven v4 and Eleven v4 Turbo

[Eleven v4 and Eleven v4 Turbo](/docs/overview/capabilities/text-to-speech/eleven-v4) are now available. Eleven v4 provides higher-quality, more expressive speech and improved voice cloning across more than 90 languages. Eleven v4 Turbo provides the same model family for real-time applications, with median inference latency of approximately 100 ms.

Use `eleven_v4` through the Text to Dialogue API for content creation and long-form audio. Use `eleven_v4_turbo` through the Text to Dialogue WebSocket for agents and interactive applications.

<YoutubeEmbed id="4QHFkK2MTcw" />

<CardGroup cols={3}>
  <Card
    title="Eleven v4 overview"
    icon="duotone book"
    href="/docs/overview/capabilities/text-to-speech/eleven-v4"
  >
    Compare the two model variants, voice cloning behavior, settings and supported languages.
  </Card>
  <Card
    title="Text to Dialogue API"
    icon="duotone code"
    href="/docs/api-reference/text-to-dialogue/convert"
  >
    Generate high-quality dialogue with Eleven v4.
  </Card>
  <Card
    title="Realtime streaming"
    icon="duotone code"
    href="/docs/eleven-api/guides/how-to/websockets/realtime-tdd"
  >
    Stream Eleven v4 Turbo dialogue over a WebSocket connection.
  </Card>
</CardGroup>

### Speech to Text

- **Transcript editing**: Batch and realtime transcription now accept a natural-language edit instruction of up to 2,000 characters. Batch responses return the result in `edited_transcript`, while realtime sessions emit an `edited_transcript` event. See the [batch](/docs/eleven-api/guides/how-to/speech-to-text/batch/transcript-editing) and [realtime](/docs/eleven-api/guides/how-to/speech-to-text/realtime/transcript-editing) guides for incompatibilities, response handling and pricing.

### ElevenAgents

- **Agent models**: Added `gpt-6-sol`, `gpt-6-luna`, `glm-52`, `deepseek-v41-flash`, `claude-opus-5` and `claude-opus-5-5` to the supported agent LLM options.
- **Agent test usage**: Test invocation summaries can include total credits and USD price. Individual test runs can include credits and a charging breakdown.

### Flows

- **Template API**: Added endpoints to list and retrieve published Flows templates, start template runs and retrieve run results. Runs can target a published version and send their terminal result to a configured webhook.

### Text to Dialogue

- **Generation continuity**: Text to Dialogue requests can use preceding or following text and request IDs to maintain continuity across generations.
- **Professional voice mode**: Text to Dialogue requests can use the Instant Voice Clone version of a Professional Voice Clone to reduce latency or improve expressiveness.

### SDK releases

#### JavaScript SDK

- [v2.70.0](https://github.com/elevenlabs/elevenlabs-js/releases/tag/v2.70.0) - Added batch and realtime Speech to Text transcript editing types, `branchId` filtering and cost fields for test invocation summaries, knowledge base auto-discovery, Text to Dialogue `usePvcAsIvc` and the `deepseek-v41-flash` agent model.

#### Python SDK

- [v2.70.0](https://github.com/elevenlabs/elevenlabs-python/releases/tag/v2.70.0) - Added batch and realtime Speech to Text transcript editing types, `branch_id` filtering and cost fields for test invocation summaries, knowledge base auto-discovery, Text to Dialogue `use_pvc_as_ivc` and the `deepseek-v41-flash` agent model.

#### ElevenLabs CLI

- [v1.4.0](https://github.com/elevenlabs/cli/releases/tag/v1.4.0) - Fixed requests with an explicit API key so they no longer also send stored OAuth credentials, which caused authentication failures. Added descriptions for the feedback command group and instructions for reporting missing capabilities to generated agent skills. Regenerated the API client with Flows template endpoints and current schema types.

### API

<Accordion title="View API changes">

## New endpoints

### Flows

- [List templates](/docs/api-reference/flows/templates/list) - `GET /v1/flows/templates` - List published templates and their runnable versions.
- [Get template](/docs/api-reference/flows/templates/get) - `GET /v1/flows/templates/{template_id}` - Retrieve a template and its runnable versions.
- [Create template run](/docs/api-reference/flows/templates/runs/create) - `POST /v1/flows/templates/{template_id}/runs` - Start a template run with input values, an optional version and an optional webhook.
- [List template runs](/docs/api-reference/flows/templates/runs/list) - `GET /v1/flows/templates/{template_id}/runs` - List template runs with cursor pagination and optional version filtering.
- [Get template run](/docs/api-reference/flows/templates/runs/get) - `GET /v1/flows/templates/{template_id}/runs/{run_id}` - Retrieve a template run and its output status.

## Updated endpoints

### Speech to Text

- [Convert speech to text](/docs/api-reference/speech-to-text/convert) - `POST /v1/speech-to-text`
  - Added optional nullable `transcript_edit` (string), which accepts a natural-language edit instruction of up to 2,000 characters.
  - Responses add optional nullable `edited_transcript`. A successful edit contains required `kind: "transcript"` and `text` fields. A failed edit contains required `kind: "error"`, `error_type: "edit_failed"` and `message` fields.
- [Realtime Speech to Text](/docs/api-reference/speech-to-text/v-1-speech-to-text-realtime) accepts `transcript_edit` (string) and emits `edited_transcript` events with required `text` and `edited_text` fields.

### ElevenAgents

- [Create agent](/docs/api-reference/agents/create), [Get agent](/docs/api-reference/agents/get) and [Update agent](/docs/api-reference/agents/update)
  - Added `gpt-6-sol`, `gpt-6-luna`, `glm-52`, `deepseek-v41-flash`, `claude-opus-5` and `claude-opus-5-5` to the `LLM` enum.
  - Alerting monitors add optional nullable `enabled` (boolean) to control whether a monitor can send notifications.
- [List conversations](/docs/api-reference/conversations/list) and [Text search](/docs/api-reference/conversations/messages/text-search)
  - `data_collection_params` and `dynamic_variable_params` add `neq` and `in` operators. Values for `in` use a pipe delimiter.
  - `evaluation_params` accepts `success`, `failure` or `unknown` as the result value.
- [Smart search](/docs/api-reference/conversations/messages/search) - `GET /v1/convai/conversations/messages/smart-search`
  - Added optional nullable `branch_id` (string).
- [List test invocations](/docs/api-reference/tests/test-invocations/list) - `GET /v1/convai/test-invocations`
  - Added optional nullable `branch_id` (string).
  - Test invocation summaries add optional nullable `credits_used` (integer) and `total_price` (number).
- [Get test invocation](/docs/api-reference/tests/test-invocations/get) - `GET /v1/convai/test-invocations/{test_invocation_id}`
  - Individual test runs add optional nullable `credits_used` (integer) and `charging` (`ConversationChargingCommonModel`).
- [Simulate conversation](/docs/api-reference/agents/simulate-conversation) and [Simulate conversation stream](/docs/api-reference/agents/simulate-conversation-stream)
  - Changed the optional `new_turns_limit` default from `10,000` to `100` and set its maximum to `100`.
- [Create crawl job](/docs/api-reference/knowledge-base/create-crawl-job) and [Get crawl job](/docs/api-reference/knowledge-base/get-crawl-job)
  - Added optional `auto_discover` (boolean, default `false`) to follow links and add newly discovered pages during auto-sync. This requires `enable_auto_sync: true`.
- [Refresh knowledge base document](/docs/api-reference/knowledge-base/refresh) - `POST /v1/convai/knowledge-base/{documentation_id}/refresh`
  - `AutoSyncInfo` adds optional `auto_discover` (boolean, default `false`) to indicate whether refreshes crawl and add newly discovered pages.
- [List MCP server tools](/docs/api-reference/mcp/list-tools) - `GET /v1/convai/mcp-servers/{mcp_server_id}/tools`
  - Responses add optional `tool_approval_statuses` (array). Each item contains required `tool_id` (string) and `state` (`up_to_date`, `needs_review` or `not_approved`), with optional approval policy and approved-definition fields.

### Text to Dialogue

- [Convert text to dialogue](/docs/api-reference/text-to-dialogue/convert), [Stream text to dialogue](/docs/api-reference/text-to-dialogue/stream), [Convert with timestamps](/docs/api-reference/text-to-dialogue/convert-with-timestamps) and [Stream with timestamps](/docs/api-reference/text-to-dialogue/stream-with-timestamps)
  - Added optional `use_pvc_as_ivc` (boolean, default `false`) to use the Instant Voice Clone version of a Professional Voice Clone.
  - Added optional nullable `previous_text` and `future_text` (strings, maximum 100 characters) and `previous_request_ids` and `next_request_ids` (string arrays, maximum 3 IDs) for generation continuity.
  - The optional `settings` object adds nullable `similarity` (number, 0–1, default `0.75`).

### Text to Speech

- [Convert text to speech](/docs/api-reference/text-to-speech/convert), [Stream text to speech](/docs/api-reference/text-to-speech/stream), [Convert with timestamps](/docs/api-reference/text-to-speech/convert-with-timestamps) and [Stream with timestamps](/docs/api-reference/text-to-speech/stream-with-timestamps)
  - The existing `use_pvc_as_ivc` field is no longer deprecated.

### Flows

- [Create image generation](/docs/api-reference/flows/image/create), [Create video generation](/docs/api-reference/flows/video/create) and [Create template run](/docs/api-reference/flows/templates/runs/create)
  - Webhook targets now require an explicit `type` discriminator (`all` or `ids`). The previous defaults were removed.

### Voice Design

- [Create voice from preview](/docs/api-reference/text-to-voice/create), [Create voice previews](/docs/api-reference/legacy/voices/create-previews), [Design a voice](/docs/api-reference/text-to-voice/design) and [Remix a voice](/docs/api-reference/text-to-voice/remix)
  - Added a `409` response when a generation for the preview is already in progress. Retry the request after receiving this response.

### Workspaces

- [Create service account API key](/docs/api-reference/service-accounts/api-keys/create) - `POST /v1/service-accounts/{service_account_user_id}/api-keys`
  - Added optional nullable `tts_concurrency_limit`, `music_concurrency_limit` and `dubbing_concurrency_limit` (integers) for enterprise service account API keys.
- [Update service account API key](/docs/api-reference/service-accounts/api-keys/update) - `PATCH /v1/service-accounts/{service_account_user_id}/api-keys/{api_key_id}`
  - Added optional `tts_concurrency_limit`, `music_concurrency_limit` and `dubbing_concurrency_limit`. Each field accepts an integer, `clear`, `no_update` or `null` and defaults to `no_update`.

### Models

- [List models](/docs/api-reference/models/list) - `GET /v1/models`
  - Deprecated required `max_characters_request_free_user` and `max_characters_request_subscribed_user`. These fields are not enforced; use `maximum_text_length_per_request` instead.

</Accordion>

> ElevenLabs provides APIs and SDKs for text to speech, voice cloning, speech to text, sound effects, voice isolator, voice changer, and conversational AI agents. Build voice-enabled applications with lifelike audio generation.