GET
/v1/convai/conversations/messages/text-search
ElevenLabs APIText Search Conversation Messages
- Base URL
- https://api.elevenlabs.io
- Auth
- xi-api-key: <ELEVENLABS_API_KEY>
- Last verified
- 2026-09-03 · upstream hash matched
Actions
Agents: curl -H "Accept: text/markdown" this URL
→ 206 tokens · Vary: Accept
→ 206 tokens · Vary: Accept
Critical gotchas
Authentication uses the non-standard xi-api-key header, not Authorization: Bearer. Sending a Bearer token returns HTTP 401.
Text-to-speech responses are raw binary audio, not JSON. Write the body to a file (--output speech.mp3) rather than parsing it.
cURL
curl -X GET 'https://api.elevenlabs.io/v1/convai/conversations/messages/text-search?text_query=example-text-query&agent_id=agent-id&visited_agent_ids=example&visited_agent_branch_ids=example&triggered_procedure_ids=example&call_successful=success&call_start_before_unix=10&call_start_after_unix=10&call_duration_min_secs=10&call_duration_max_secs=10&rating_max=10&rating_min=10&has_feedback_comment=true&user_id=user-id&evaluation_params=example&data_collection_params=example&tool_names=example&tool_names_successful=example&tool_names_errored=example&include_invalid_tool_calls=false&main_languages=example&exclude_statuses=example&termination_reasons=example&page_size=20&summary_mode=exclude&conversation_initiation_source=unknown&text_only=true&conversation_product_type=agents&branch_id=branch-id&version_id=version-id&topic_ids=example&sort_by=search_score&cursor=example-cursor' \ -H "xi-api-key: $ELEVENLABS_API_KEY"Get a free ElevenLabs API key → sponsored
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| text_query | query | string | Yes | The search query text for full-text and fuzzy matching |
| agent_id | query | string | No | Agent id (agent_…) or speech engine external id (seng_), resolved to the same underlying resource. |
| visited_agent_ids | query | string[] | No | Filter conversations where any of these agents participated. Can not exceed 50 values. |
| visited_agent_branch_ids | query | string[] | No | Filter conversations where any of these agent branches participated. Can not exceed 50 values. |
| triggered_procedure_ids | query | string[] | No | Filter conversations where any of these procedures were triggered. Can not exceed 50 values. |
| call_successful | query | string | No | The result of the success evaluation |
| call_start_before_unix | query | integer | No | Unix timestamp (in seconds) to filter conversations up to this start date. |
| call_start_after_unix | query | integer | No | Unix timestamp (in seconds) to filter conversations after to this start date. |
| call_duration_min_secs | query | integer | No | Minimum call duration in seconds. |
| call_duration_max_secs | query | integer | No | Maximum call duration in seconds. |
| rating_max | query | integer | No | Maximum overall rating (1-5). |
| rating_min | query | integer | No | Minimum overall rating (1-5). |
| has_feedback_comment | query | boolean | No | Filter conversations with user feedback comments. |
| user_id | query | string | No | Filter conversations by the user ID who initiated them. |
| evaluation_params | query | string[] | No | Evaluation filters. Repeat param. Format: criteria_id:result. Example: eval=value_framing:success |
| data_collection_params | query | string[] | No | Data collection filters. Repeat param. Format: id:op:value where op is one of eq|neq|gt|gte|lt|lte|in|exists|missing. For in, pipe-delimit values. |
| tool_names | query | string[] | No | Filter conversations by tool names used during the call. |
| tool_names_successful | query | string[] | No | Filter conversations by tool names that had successful calls. |
| tool_names_errored | query | string[] | No | Filter conversations by tool names that had errored calls. |
| include_invalid_tool_calls | query | boolean | No | Also match tool calls that never ran. |
| main_languages | query | string[] | No | Filter conversations by detected main language (language code). |
| exclude_statuses | query | string[] | No | Exclude conversations with the given statuses. Useful for hiding in-progress / processing conversations from list views. |
| termination_reasons | query | string[] | No | Filter conversations by their stored termination_reason (metadata.termination_reason). Repeat param to match any of several. |
| page_size | query | integer | No | Number of results per page. Max 50. |
| summary_mode | query | string | No | Whether to include transcript summaries in the response. |
| conversation_initiation_source | query | string | No | query parameter conversation_initiation_source. |
| text_only | query | boolean | No | query parameter text_only. |
| conversation_product_type | query | string | No | Restrict results to a single conversation product surface. |
| branch_id | query | string | No | Filter conversations by branch ID. |
| version_id | query | string | No | Filter conversations by version ID. |
| topic_ids | query | string[] | No | Filter conversations by topic IDs assigned during topic discovery. |
| sort_by | query | string | No | Sort order for search results. 'search_score' sorts by search score, 'created_at' sorts by conversation start time. |
| cursor | query | string | No | Used for fetching next page. Cursor is returned in the response. |
Response 200 OK
{
"meta": {
"total": 0,
"page": 0,
"page_size": 0
},
"results": [
{
"conversation_id": "string",
"agent_id": "string",
"agent_name": "string",
"transcript_index": 0,
"chunk_text": "string",
"chunk_highlights": [
{
"value": "string",
"is_hit": true
}
],
"score": 0,
"conversation_start_time_unix_secs": 0
}
],
"next_cursor": "string",
"has_more": true
}