feat(audio): add custom OpenAI-compatible TTS/ASR provider support (#357)#409

Merged

cosarah merged 26 commits intomainfrom

worktree-357-custom-tts-asr-provider

Apr 12, 2026

Commits on Apr 12, 2026

feat(audio): extend TTSProviderId/ASRProviderId to support custom providers
wyuc
committed
feat(audio): update provider registry helpers for custom provider support
wyuc
committed
feat(audio): route custom providers to OpenAI-compatible implementations

wyuc
and
claude
committed
feat(audio): add custom TTS/ASR provider CRUD to settings store
wyuc
committed
feat(audio): add dialog for creating custom audio providers
wyuc
committed
feat(audio): add custom provider UI with voice management and delete

wyuc
and
claude
committed
feat(i18n): add custom audio provider translation strings
wyuc
committed
fix(audio): resolve remaining type errors for custom provider support

wyuc
and
claude
committed
fix(audio): allow custom provider IDs to reach switch default branch

wyuc
and
claude
committed
feat(audio): add default model field to dialog and polish voice list UI

wyuc
and
claude
committed
fix(audio): fallback to customDefaultBaseUrl when testing custom TTS provider

wyuc
and
claude
committed
fix(audio): auto-select first voice when adding to custom TTS provider

wyuc
and
claude
committed
fix(audio): fallback to customDefaultBaseUrl in all TTS call sites

wyuc
and
claude
committed
fix(audio): include custom provider voices in agent voice picker

wyuc
and
claude
committed
fix(audio): fallback to customDefaultBaseUrl in all ASR call sites

wyuc
and
claude
committed
feat(audio): add model list management for custom ASR providers

wyuc
and
claude
committed
fix(audio): seed customModels from defaultModel and fix ASR model label

wyuc
and
claude
committed
fix(audio): separate ASR model management from creation dialog

wyuc
and
claude
committed
fix(audio): simplify ASR model management — first model is default

wyuc
and
claude
committed
fix(audio): add loading state and detailed errors to ASR test

wyuc
and
claude
committed
fix(i18n): use correct key for processing label in ASR test button
wyuc
committed
fix(audio): polish custom provider UX and type safety

wyuc
and
claude
committed
fix(test): add isCustomTTSProvider/isCustomASRProvider to audio types mock

wyuc
and
claude
committed
fix(audio): add customDefaultBaseUrl fallback in all TTS/ASR call sites

wyuc
and
claude
committed
style: format generation-preview/page.tsx

wyuc
and
claude
committed
fix: remove unused ttsSpeedRange variable in media-popover

wyuc
and
claude
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(audio): add custom OpenAI-compatible TTS/ASR provider support (#357)#409

feat(audio): add custom OpenAI-compatible TTS/ASR provider support (#357)#409
cosarah merged 26 commits intomainfrom
worktree-357-custom-tts-asr-provider

Commits on Apr 12, 2026

feat(audio): extend TTSProviderId/ASRProviderId to support custom providers

feat(audio): update provider registry helpers for custom provider support

feat(audio): route custom providers to OpenAI-compatible implementations

feat(audio): add custom TTS/ASR provider CRUD to settings store

feat(audio): add dialog for creating custom audio providers

feat(audio): add custom provider UI with voice management and delete

feat(i18n): add custom audio provider translation strings

fix(audio): resolve remaining type errors for custom provider support

fix(audio): allow custom provider IDs to reach switch default branch

feat(audio): add default model field to dialog and polish voice list UI

fix(audio): fallback to customDefaultBaseUrl when testing custom TTS provider

fix(audio): auto-select first voice when adding to custom TTS provider

fix(audio): fallback to customDefaultBaseUrl in all TTS call sites

fix(audio): include custom provider voices in agent voice picker

fix(audio): fallback to customDefaultBaseUrl in all ASR call sites

feat(audio): add model list management for custom ASR providers

fix(audio): seed customModels from defaultModel and fix ASR model label

fix(audio): separate ASR model management from creation dialog

fix(audio): simplify ASR model management — first model is default

fix(audio): add loading state and detailed errors to ASR test

fix(i18n): use correct key for processing label in ASR test button

fix(audio): polish custom provider UX and type safety

fix(test): add isCustomTTSProvider/isCustomASRProvider to audio types mock

fix(audio): add customDefaultBaseUrl fallback in all TTS/ASR call sites

style: format generation-preview/page.tsx

fix: remove unused ttsSpeedRange variable in media-popover