aid: lmnt
name: LMNT
type: Index
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
url: https://raw.githubusercontent.com/api-evangelist/lmnt/refs/heads/main/apis.yml
created: 2026-06-12
modified: 2026-06-12
specificationVersion: "0.19"
description: >
  LMNT is a text-to-speech API platform delivering ultra-low latency voice synthesis
  with streaming audio output designed for real-time conversational AI applications.
  The platform provides a Speech API for standard text-to-speech generation and a
  Speech Sessions API for WebSocket-based real-time streaming integrated with LLM
  pipelines, achieving latency under 300 milliseconds. LMNT supports 31 languages
  and offers voice cloning from as little as five seconds of audio, with its Blizzard 2
  model optimized for accuracy, expressiveness, and pronunciation. Authentication uses
  API keys managed via the app dashboard, with SDKs available for Python, TypeScript,
  and Go.
tags:
  - text-to-speech
  - voice synthesis
  - voice cloning
  - audio streaming
  - conversational AI
  - low latency
  - real-time audio
apis:
  - name: LMNT Speech API
    description: >
      REST API for converting text to speech audio, supporting streaming binary
      audio output, 31 languages, voice cloning, word timestamps, and configurable
      expressiveness parameters.
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    humanURL: https://docs.lmnt.com/api-reference/speech/synthesize-speech-bytes
    baseURL: https://api.lmnt.com
    tags:
      - text-to-speech
      - audio
      - streaming
    properties:
      - type: Documentation
        url: https://docs.lmnt.com/api-reference/speech/synthesize-speech-bytes
      - type: OpenAPI
        url: openapi/lmnt-speech-openapi.yml
  - name: LMNT Speech Sessions API
    description: >
      WebSocket-based real-time speech generation API for streaming LLM text output
      to synthesized audio with reset-latency support for conversational AI
      applications requiring interrupt handling.
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    humanURL: https://docs.lmnt.com/api-reference/speech/generate-speech-session
    baseURL: wss://api.lmnt.com
    tags:
      - speech sessions
      - websocket
      - real-time
      - streaming
    properties:
      - type: Documentation
        url: https://docs.lmnt.com/api-reference/speech/generate-speech-session
common:
  - type: Website
    url: https://www.lmnt.com/
  - type: Documentation
    url: https://docs.lmnt.com/
  - type: GitHubOrganization
    url: https://github.com/lmnt-com
  - type: LinkedIn
    url: https://www.linkedin.com/company/lmnt
  - type: X
    url: https://x.com/lmnt_com
  - type: Pricing
    url: https://www.lmnt.com/pricing
  - type: Changelog
    url: https://docs.lmnt.com/changelog/overview
  - type: Plans
    url: plans/lmnt-plans-pricing.yml
  - type: RateLimits
    url: rate-limits/lmnt-rate-limits.yml
  - type: FinOps
    url: finops/lmnt-finops.yml
  - type: OpenAPI
    url: openapi/lmnt-speech-openapi.yml
  - type: JSONSchema
    url: json-schema/lmnt-voice-schema.json
  - type: JSONSchema
    url: json-schema/lmnt-speech-synthesis-schema.json
  - type: JSONLDContext
    url: json-ld/lmnt-context.jsonld
  - type: Vocabulary
    url: vocabulary/lmnt-vocabulary.yml
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com