generated: '2026-09-17' method: derived source: openapi/elevenlabs-openapi.json (OpenAPI 3.1.0, 301 paths, 390 operations, 1,520 component schemas — the provider's own published contract, fetched from https://api.elevenlabs.io/openapi.json on 2026-09-17) description: | Entity-relationship graph for the ElevenLabs API, derived from path templates and the id-reference fields in components.schemas. Relationships are read from foreign-key-shaped fields (`*_id`) and from path nesting — the contract uses $ref for schema composition rather than for links, so identifier fields are the only relationship signal available. Cardinalities marked confidence: medium are inferred from path nesting; nothing here is asserted from prose. id_convention: style: opaque string identifiers, no type prefix note: | Unlike Stripe-shaped APIs, ElevenLabs identifiers carry no type prefix — a voice_id (e.g. JBFqnCBsd6RMkjVDRZzb) and an agent_id (e.g. agent_7101k5zvyjhmfg983brhmhkd98n6) do not share a scheme, and most ids are bare opaque strings. Agent identifiers observed in the widget docs DO carry an `agent_` prefix; that is the only prefixed family seen. no_published_id_reference: the docs publish no object/id reference page, so prefix rules cannot be stated beyond what appears in examples. domains: - domain: ElevenAgents (Conversational AI) root_path: /v1/convai operations: 174 entities: - agents - conversations - knowledge-base - tools - mcp-servers - phone-numbers - batch-calling - agent-testing - triage-tickets - secrets - environment-variables - whatsapp-accounts - tags - settings - domain: ElevenCreative — Studio root_path: /v1/studio entities: - projects - chapters - snapshots - domain: ElevenCreative — Dubbing / Productions root_path: /v1/dubbing, /v1/productions entities: - dubbing projects - dubbing resources - speakers - segments - production orders - domain: Voices root_path: /v1/voices entities: - voices - samples - pvc voices - voice settings - voice library - domain: Generation root_path: /v1/text-to-speech, /v1/speech-to-text, /v1/speech-to-speech, /v1/music, /v1/sound-generation, /v1/audio-isolation, /v1/text-to-dialogue entities: - generations - history items - music finetunes - domain: Workspace / administration root_path: /v1/workspace, /v1/service-accounts, /v1/usage entities: - workspace - groups - webhooks - auth-connections - api-keys - usage entities: - name: workspace key: null description: the account boundary. Every other entity is scoped to exactly one workspace; the workspace itself is addressed implicitly by the credential, not by an id in the path. - name: agent key: agent_id path: /v1/convai/agents/{agent_id} operations: 40 description: an ElevenAgents conversational agent — prompt, voice, language, tools, knowledge base, widget and platform settings. - name: agent_branch key: branch_id path: /v1/convai/agents/{agent_id}/branches/{branch_id} description: a versioned branch of an agent's configuration. With version_id this is the reversal path for a bad agent change. - name: conversation key: conversation_id path: /v1/convai/conversations/{conversation_id} description: one completed or in-flight conversation with an agent, carrying the transcript, metrics and analysis. - name: knowledge_base_document key: documentation_id path: /v1/convai/knowledge-base/{documentation_id} operations: 31 description: a document (file, URL or text) attached to an agent's RAG knowledge base. - name: tool key: tool_id path: /v1/convai/tools/{tool_id} description: a webhook or client tool an agent may call. - name: mcp_server key: mcp_server_id path: /v1/convai/mcp-servers/{mcp_server_id} description: an EXTERNAL MCP server an ElevenLabs agent connects to as a client — the inverse direction from the hosted MCP server in mcp/elevenlabs-mcp.yml. - name: phone_number key: phone_number_id path: /v1/convai/phone-numbers/{phone_number_id} - name: batch_call key: batch_id path: /v1/convai/batch-calling/{batch_id} - name: voice key: voice_id path: /v1/voices/{voice_id} description: a voice — premade, instant clone (IVC), professional clone (PVC) or designed. - name: sample key: sample_id path: /v1/voices/{voice_id}/samples/{sample_id} description: an audio sample belonging to a cloned voice. - name: project key: project_id path: /v1/studio/projects/{project_id} description: a Studio project (audiobook, podcast, long-form voiceover). - name: chapter key: chapter_id path: /v1/studio/projects/{project_id}/chapters/{chapter_id} - name: project_snapshot key: project_snapshot_id path: /v1/studio/projects/{project_id}/snapshots/{project_snapshot_id} - name: dubbing_project key: dubbing_id path: /v1/dubbing/{dubbing_id} - name: dubbing_speaker key: speaker_id path: /v1/dubbing/resource/{dubbing_id}/speaker/{speaker_id} deprecated: true - name: dubbing_segment key: segment_id path: /v1/dubbing/resource/{dubbing_id}/segment/{segment_id} deprecated: true - name: production_order key: order_id path: /v1/productions/orders/{order_id} - name: pronunciation_dictionary key: pronunciation_dictionary_id path: /v1/pronunciation-dictionaries/{pronunciation_dictionary_id} - name: history_item key: history_item_id path: /v1/history/{history_item_id} description: a past generation, with its audio and the request that produced it. - name: model key: model_id description: a generation model (eleven_v3, eleven_multilingual_v2, scribe_v2, music_v2_5 ...). Referenced by 57 schemas; addressed as a list-only resource. - name: webhook key: webhook_id path: /v1/workspace/webhooks/{webhook_id} - name: api_key key: key_id path: /v1/service-accounts/{service_account_id}/api-keys/{key_id} relationships: - from: workspace to: agent type: has_many via: implicit (credential scope) confidence: high - from: workspace to: webhook type: has_many via: /v1/workspace/webhooks confidence: high - from: agent to: conversation type: has_many via: agent_id confidence: high - from: agent to: agent_branch type: has_many via: /v1/convai/agents/{agent_id}/branches confidence: high - from: agent to: knowledge_base_document type: has_many via: knowledge base attachment on the agent configuration confidence: medium - from: agent to: tool type: has_many via: tool_ids confidence: high - from: agent to: mcp_server type: has_many via: native_mcp_server_ids confidence: high - from: agent to: voice type: has_one via: voice_id in the TTS configuration confidence: high - from: agent to: phone_number type: has_many via: phone_number_id confidence: medium - from: batch_call to: agent type: belongs_to via: agent_id confidence: high - from: conversation to: voice type: has_one via: the agent's voice at call time confidence: medium - from: voice to: sample type: has_many via: /v1/voices/{voice_id}/samples confidence: high - from: project to: chapter type: has_many via: /v1/studio/projects/{project_id}/chapters confidence: high - from: project to: project_snapshot type: has_many via: /v1/studio/projects/{project_id}/snapshots confidence: high - from: chapter to: chapter_snapshot type: has_many via: chapter_snapshot_id confidence: high - from: project to: voice type: has_many via: voice_id per chapter/paragraph confidence: medium - from: dubbing_project to: dubbing_speaker type: has_many via: /v1/dubbing/resource/{dubbing_id}/speaker confidence: high - from: dubbing_speaker to: dubbing_segment type: has_many via: /v1/dubbing/resource/{dubbing_id}/speaker/{speaker_id}/segment confidence: high - from: history_item to: voice type: belongs_to via: voice_id confidence: high - from: history_item to: model type: belongs_to via: model_id confidence: high - from: generation to: model type: belongs_to via: model_id confidence: high - from: generation to: pronunciation_dictionary type: has_many via: pronunciation_dictionary_locators confidence: medium - from: folder to: folder type: belongs_to via: parent_folder_id / folder_parent_id confidence: high note: self-referential folder tree used by Studio and the agents workspace. counts: component_schemas: 1520 paths: 301 operations: 390 entities_modelled: 24 relationships_modelled: 23