feat(ai-groq): transcription#649
Open
joksas wants to merge 5 commits into
Open
Conversation
Contributor
📝 WalkthroughWalkthroughAdds Groq transcription support: model metadata and provider options, a GroqTranscriptionAdapter that accepts URLs or local audio, factories (explicit/API-key env), package re-exports, a changeset, E2E test wiring, and updated docs. ChangesGroq Transcription Support
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@packages/ai-groq/src/adapters/transcription.ts`:
- Around line 161-170: The adapter currently forwards
TranscriptionOptions.responseFormat (effectiveFormat) including 'srt'/'vtt' to
form.append('response_format', effectiveFormat) and later calls response.json()
for non-verbose_json/text branches; change this to explicitly reject unsupported
formats for Groq: add a check after computing effectiveFormat (the variable
named effectiveFormat in transcription.ts) and if effectiveFormat === 'srt' ||
effectiveFormat === 'vtt' throw a clear Error like "responseFormat 'srt'/'vtt'
not supported by Groq" (or alternatively map them to a supported value such as
'text' or 'json' if you prefer), and ensure form.append('response_format', ...)
and the subsequent response handling (the branches that call response.text() or
response.json(), e.g., the block that currently uses response.json() and
generateId(this.name)) only run for supported formats
('json','text','verbose_json').
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: f2e8fc62-7779-4e78-9f21-0e407ea9246f
📒 Files selected for processing (7)
.changeset/feat-groq-ai-transcription.mdpackages/ai-groq/src/adapters/transcription.tspackages/ai-groq/src/audio/transcription-provider-options.tspackages/ai-groq/src/index.tspackages/ai-groq/src/model-meta.tstesting/e2e/src/lib/feature-support.tstesting/e2e/src/lib/media-providers.ts
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/media/transcription.md`:
- Around line 99-102: The comment incorrectly states that segments include
word-level timing; update the documentation comment to clarify that verbose_json
is the default, segment timestamps live on result.segments (segment-level
start/end) and word-level timing is exposed separately on result.words
(top-level), and adjust any example text around result.segments and result.words
to reflect that contract (refer to symbols result.segments, result.words, and
verbose_json).
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 2f7c9fa6-04a5-47a4-a758-3fe3aee26176
📒 Files selected for processing (2)
docs/adapters/groq.mddocs/media/transcription.md
Comment on lines
+99
to
+102
| // verbose_json is the default — segments include word-level timing when requested | ||
| for (const segment of result.segments ?? []) { | ||
| console.log(`[${segment.start}s → ${segment.end}s] ${segment.text}`) | ||
| } |
Contributor
There was a problem hiding this comment.
Correct the word-timestamp shape in the Groq example comment.
The note says segments include word-level timing, but the documented contract exposes word timing via result.words (top-level), while segments are segment-level timestamps.
Suggested doc tweak
-// verbose_json is the default — segments include word-level timing when requested
+// verbose_json is the default — segment timestamps are in `result.segments`
+// and word-level timing (if requested/supported) is in `result.words`
for (const segment of result.segments ?? []) {
console.log(`[${segment.start}s → ${segment.end}s] ${segment.text}`)
}🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/media/transcription.md` around lines 99 - 102, The comment incorrectly
states that segments include word-level timing; update the documentation comment
to clarify that verbose_json is the default, segment timestamps live on
result.segments (segment-level start/end) and word-level timing is exposed
separately on result.words (top-level), and adjust any example text around
result.segments and result.words to reflect that contract (refer to symbols
result.segments, result.words, and verbose_json).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
🎯 Changes
Adds Groq as a transcription provider. Groq's API is mostly OpenAI SDK-compatible, but its transcription endpoint additionally accepts HTTP URLs as input, so this is implemented as a custom integration rather than going through the SDK.
✅ Checklist
pnpm run test:pr.🚀 Release Impact
Summary by CodeRabbit
New Features
Documentation
Tests