Skip to content

docs(openrouter-stt): sync with openrouter-web changes #75

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
devin-ai-integration wants to merge 1 commit into
base: main
Choose a base branch
from
Open

docs(openrouter-stt): sync with openrouter-web changes #75

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 19 additions & 0 deletions skills/openrouter-stt/SKILL.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,6 +98,8 @@ Models are provider-namespaced — use the full slug (`google/chirp-3`, `openai/
| `input_audio.format` | yes | `wav`, `mp3`, `flac`, `m4a`, `ogg`, `webm`, or `aac`. Must match the actual bytes. Support varies by provider. |
| `language` | no | ISO-639-1 code (`en`, `ja`, `fr`). Auto-detected if omitted. |
| `temperature` | no | 0–1. Lower is more deterministic. |
| `response_format` | no | `json` (default) or `verbose_json` — see below. `text`, `srt`, and `vtt` are rejected with a 400. |
| `timestamp_granularities` | no | Array of `"segment"` and/or `"word"`. Only meaningful with `response_format: "verbose_json"`. |
| `provider` | no | Provider passthrough — see below. |

### Picking an audio format
Expand All @@ -108,6 +110,23 @@ Models are provider-namespaced — use the full slug (`google/chirp-3`, `openai/

The `format` field must match the actual container/codec of the bytes. A file saved as `.wav` that is actually mp3 will be rejected or mis-decoded. When in doubt, confirm with `ffprobe <file>`.

## Timestamps with `verbose_json`

Set `response_format: "verbose_json"` to additionally get `task`, `language`, `duration` (seconds), and timestamped `segments` in the response. Pass `timestamp_granularities: ["word"]` to also receive word-level timestamps in a `words` array (`segment` is the provider default).

```json
{
"model": "openai/whisper-large-v3",
"input_audio": { "data": "UklGRiQA...", "format": "wav" },
"response_format": "verbose_json",
"timestamp_granularities": ["word"]
}
```

Each segment carries `id`, `start`, `end` (seconds), `text`, and may include Whisper-style extras (`tokens`, `avg_logprob`, `compression_ratio`, `no_speech_prob`). Each word carries `word`, `start`, `end`.

`verbose_json` is only available on OpenAI-compatible providers (OpenAI, Groq, Together) — other providers, and a few specific models (e.g. OpenAI's gpt-4o transcribe models), reject it with a 400. Default `json` keeps working everywhere.

## Provider-specific options

Provider passthrough goes under `provider.options.<slug>` and is only forwarded when that provider handles the request. Example — Groq's `prompt` for vocabulary hinting:
Expand Down