v0.9.1

Release 0.9.1

Overview

Internal dependency update: speech web components now build against @corti/sdk 6.0.0. No component API changes — this release lets consumers consume the components on the new SDK major line.

❗ Breaking changes

• None. Component properties, events, and element tags are unchanged.

Changes

• chore: upgrade @corti/sdk from 5.0.0 to 6.0.0 (workspace root, dictation, ambient).

v0.9.1-rc.2

Release 0.9.1

Overview

Internal dependency update: speech web components now build against @corti/sdk 6.0.0. No component API changes — this release lets consumers consume the components on the new SDK major line.

❗ Breaking changes

• None. Component properties, events, and element tags are unchanged.

Changes

• chore: upgrade @corti/sdk from 5.0.0 to 6.0.0 (workspace root, dictation, ambient).

v0.9.1-rc

Release 0.9.1

Overview

Internal dependency update: speech web components now build against @corti/sdk 6.0.0. No component API changes — this release lets consumers (console, examples) consume the components on the new SDK major line.

❗ Breaking changes

• None. Component properties, events, and element tags are unchanged.

Changes

• chore: upgrade @corti/sdk from 5.0.0 to 6.0.0 (workspace root, dictation, ambient).

v0.9.0

Release

Overview

Speech web components depend on @corti/sdk 5.0.0, send component identity on every request via x-corti-analytics (web_component and web_component_version), and load dictation/ambient language lists from GET /languages/ when authenticated.

❗ Breaking changes

• None.

Changes

• chore: upgrade @corti/sdk from 4.0.0 to 5.0.0.
• feat: send web_component and web_component_version on x-corti-analytics for HTTP and WebSocket traffic from @corti/dictation-web and @corti/ambient-web. Extra keys can be passed via the analytics property.
• feat: expose a readonly version property on dictation and ambient elements.
• feat: load available languages from GET /languages/ (transcribe for dictation, streams for ambient) and fall back to the region list when unauthenticated or the request fails.

v0.8.0

Overview

@corti/dictation-web and @corti/ambient-web now depend on stable @corti/sdk@4.0.0. Ambient default and virtual-mode stream configuration uses the SDK v4 diarize field instead of the removed isDiarization option.

❗ Breaking changes

• SDK v4 required: Both packages now depend on @corti/sdk@4.0.0. Upgrade your app’s SDK dependency when upgrading these components.

  • Migration: Bump @corti/sdk to 4.0.0 (or newer) alongside @corti/dictation-web / @corti/ambient-web.
  • Example:
    • Before: "@corti/sdk": "3.0.0", "@corti/ambient-web": "0.7.0"
    • After: "@corti/sdk": "4.0.0", "@corti/ambient-web": "0.8.0"

• Ambient isDiarization removed: Default ambient config and virtual-mode toggling now set transcription.diarize per SDK v4.

  • Migration: If you pass a custom ambientConfig (or merge into Corti.StreamConfig), rename transcription.isDiarization to transcription.diarize.
  • Example:
    • Before: transcription: { isDiarization: true, primaryLanguage: "en" }
    • After: transcription: { diarize: true, primaryLanguage: "en" }

Changes

• chore(ambient): use transcription.diarize in DEFAULT_AMBIENT_CONFIG and virtual-mode config helper.
• chore(deps): bump @corti/sdk from 3.0.0 to 4.0.0 in @corti/dictation-web and @corti/ambient-web.

v0.8.0-rc.1

Release 0.8.0

Overview

@corti/dictation-web and @corti/ambient-web now depend on stable @corti/sdk@4.0.0. Ambient default and virtual-mode stream configuration uses the SDK v4 diarize field instead of the removed isDiarization option.

❗ Breaking changes

• SDK v4 required: Both packages now depend on @corti/sdk@4.0.0. Upgrade your app’s SDK dependency when upgrading these components.

  • Migration: Bump @corti/sdk to 4.0.0 (or newer) alongside @corti/dictation-web / @corti/ambient-web.
  • Example:
    • Before: "@corti/sdk": "3.0.0", "@corti/ambient-web": "0.7.0"
    • After: "@corti/sdk": "4.0.0", "@corti/ambient-web": "0.8.0"

• Ambient isDiarization removed: Default ambient config and virtual-mode toggling now set transcription.diarize per SDK v4.

  • Migration: If you pass a custom ambientConfig (or merge into Corti.StreamConfig), rename transcription.isDiarization to transcription.diarize.
  • Example:
    • Before: transcription: { isDiarization: true, primaryLanguage: "en" }
    • After: transcription: { diarize: true, primaryLanguage: "en" }

Changes

• chore(ambient): use transcription.diarize in DEFAULT_AMBIENT_CONFIG and virtual-mode config helper.
• chore(deps): bump @corti/sdk from 3.0.0 to 4.0.0 in @corti/dictation-web and @corti/ambient-web.

v0.8.0-rc

chore: update corti sdk to 4.0.0-rc

v0.7.0

Overview

This release splits the repository into a pnpm monorepo with shared core/ internals and two published packages: @corti/dictation-web (Transcribe) and the new @corti/ambient-web (Streams). Dictation custom elements and their public API are unchanged for existing integrators; ambient adds a full all-in-one and modular component set for multi-speaker real-time streaming.

❗ Breaking changes

• None.

Changes

@corti/ambient-web (new)

• feat(ambient): add @corti/ambient-web — web components for real-time multi-speaker ambient streaming. See the Ambient Web Component docs.

@corti/dictation-web

• feat(dictation): dispatch audioEvent on <corti-dictation> and <dictation-recording-button>; export AudioEventEventDetail for typing event handlers.
• chore(dictation): internal monorepo refactor — no changes required for existing <corti-dictation> / <dictation-*> integrations.

v0.7.0-ambient.16

Packages

• @corti/dictation-web@0.7.0-ambient.16
• @corti/ambient-web@0.7.0-ambient.16

Install: npm install @corti/dictation-web@ambient @corti/ambient-web@ambient

Changes

• Drop @corti/core-web as a workspace/npm package; shared code lives under core/ with @core/* imports only at build time
• Published tarballs are self-contained: runtime bundle.js, types via index.d.ts + vendored dist/core/
• safeCustomElement on all custom elements (skips duplicate registration when dictation + ambient load on one page)

Note

v0.7.0-ambient.15 tag failed CI (lint) and was not published to npm.

v0.7.0-ambient.10

fix(ci): set package version from dictation/ and ambient/ directories

pnpm version at the workspace root does not honor --dir; run it inside each
package folder instead.