[Integrations](https://spiral.ai/integrations) › [Factory Droid](https://spiral.ai/integrations/factory-droid) › [Cohere Transcribe](https://spiral.ai/integrations/factory-droid/cohere-transcribe)

# How to Use Cohere Transcribe in Factory Droid

Cohere Transcribe turns recorded speech into text for searchable interviews, training materials, and voice notes. It supports 14 languages, works best with one specified language, and supplies neither timestamps nor speaker labels.

|  |  |
| --- | --- |
| Model | Cohere Transcribe |
| Available through | fal |
| Category | Audio |
| Price | Up to $0.005382 per run at the default inputs for 60 seconds of input. |
| Returns | json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Cohere Transcribe in Factory Droid

- Use Cohere Transcribe to transcribe the English bug report in recordings/checkout-bug.wav with punctuation. Compare the reported steps with the checkout code and draft an issue separating the user's claims from behavior you can verify.
- Use Cohere Transcribe on recordings/architecture-walkthrough.wav, specifying English and keeping punctuation. Compare the transcript with the repository and propose documentation corrections for my review.
- Use Cohere Transcribe to transcribe recordings/customer-interview.wav as Spanish with punctuation. Then draft English product notes, keeping the Spanish transcript beside each supporting excerpt.
- Use Cohere Transcribe on the English narration in recordings/reproduction.wav with punctuation. Read the relevant project files and draft a reproduction checklist without inventing missing steps.
- Use Cohere Transcribe to transcribe recordings/dictation.wav as English with punctuation disabled. Save the lowercase transcript as a text fixture for the project's normalization examples.

## How to set up Cohere Transcribe in Factory Droid

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Add the server**: Droid saves it to your user config, `~/.factory/mcp.json`, so it works in every project.

```bash
droid mcp add spiral https://api.spiral.ai/mcp --type http
```

2. **Sign in**: Start `droid`, type `/mcp`, select spiral and authenticate it. Sign in to Spiral in the browser and click Authorize. No API key.

3. **Check it**: Spiral shows as connected. needs authentication means the sign-in did not finish: run `/mcp` again.

```bash
droid mcp list
```

### Use the Spiral skill

1. **Send Factory Droid this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run Cohere Transcribe**: One request; the result comes back by polling or a webhook.

```bash
curl https://api.spiral.ai/v1/runs \
  -H "Authorization: Bearer $SPIRAL_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
    "provider": "fal",
    "model": "fal-ai/cohere-transcribe",
    "input": {
      "audio_url": "https://storage.googleapis.com/falserverless/model_tests/whisper/dinner_conversation.mp3"
    }
  }'
```

Then paste one of the prompts above: Factory Droid finds Cohere Transcribe on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/factory-droid

## Cohere Transcribe tools and prices

- [Cohere Transcribe](https://spiral.ai/tools/fal/fal-ai/cohere-transcribe) (fal): $0.0000897 per second.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| language | en, fr, de, it, es, pt, el, nl, pl, zh, ja, ko, vi, ar | en | Language of the audio (ISO 639-1 code). Supported: en, fr, de, it, es, pt, el, nl, pl, zh, ja, ko, vi, ar. |
| punctuation |  | true | Whether to include punctuation and capitalization in the output. |
| max_new_tokens | 1 to 1014 | 256 | Maximum number of tokens to generate per audio chunk. |

## What can you do with Cohere Transcribe in Factory Droid?

Cohere Transcribe makes spoken bug reports, walkthroughs, and technical interviews available as text inside a coding task. The coding agent can compare that transcript with project files and draft changes or issues grounded in what was recorded.

### Searchable interview transcripts

Transcribe interviews conducted in one supported language, then use the text to find passages for editorial review. Check selected quotations against the recording before publication.

### Training recording archives

Turn spoken training sessions into searchable reference text. An agent can then organize the transcript by topic and draft a reading guide.

### Voice notes for writing

Convert dictated ideas into text with punctuation. Have an agent draft an outline afterward while retaining the original transcript for comparison.

### Recorded call review

Create transcripts that reviewers can search for customer questions or recurring complaints. Add separate speaker identification if the review requires attribution.

### Multilingual audio archives

Transcribe recordings in any of the 14 supported languages, using the correct language for each recording. Keep the original-language text available for search and review.

### Spoken bug reports

Turn narrated reproduction steps into a transcript. A coding agent can compare that text with the project and draft an issue that distinguishes reported behavior from verified behavior.

### Meeting notes from recordings

Transcribe a meeting, then have an agent draft decisions and follow-up questions from the text. Review attribution separately because the transcript does not identify speakers.

## Overview

- [fal](https://spiral.ai/providers/fal)
- [All Factory Droid guides](https://spiral.ai/integrations/factory-droid)

## Cohere Transcribe in other agents

- [Amp](https://spiral.ai/integrations/amp/cohere-transcribe)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/cohere-transcribe)
- [Claude](https://spiral.ai/integrations/claude-desktop/cohere-transcribe)
- [Claude Code](https://spiral.ai/integrations/claude-code/cohere-transcribe)
- [Cline](https://spiral.ai/integrations/cline/cohere-transcribe)
- [Codex](https://spiral.ai/integrations/codex/cohere-transcribe)
- [Cursor](https://spiral.ai/integrations/cursor/cohere-transcribe)
- [Devin](https://spiral.ai/integrations/devin/cohere-transcribe)
- [dots](https://spiral.ai/integrations/dots/cohere-transcribe)
- [Gemini](https://spiral.ai/integrations/gemini/cohere-transcribe)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/cohere-transcribe)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/cohere-transcribe)
- [goose](https://spiral.ai/integrations/goose/cohere-transcribe)
- [Grok](https://spiral.ai/integrations/grok/cohere-transcribe)
- [Grok Build](https://spiral.ai/integrations/grok-build/cohere-transcribe)
- [Hermes](https://spiral.ai/integrations/hermes/cohere-transcribe)
- [Instinct](https://spiral.ai/integrations/instinct/cohere-transcribe)
- [Manus](https://spiral.ai/integrations/manus/cohere-transcribe)
- [Muse](https://spiral.ai/integrations/muse/cohere-transcribe)
- [Muse Code](https://spiral.ai/integrations/muse-code/cohere-transcribe)
- [n8n](https://spiral.ai/integrations/n8n/cohere-transcribe)
- [Notion](https://spiral.ai/integrations/notion/cohere-transcribe)
- [OpenClaw](https://spiral.ai/integrations/openclaw/cohere-transcribe)
- [opencode](https://spiral.ai/integrations/opencode/cohere-transcribe)
- [Perplexity](https://spiral.ai/integrations/perplexity/cohere-transcribe)
- [Raycast](https://spiral.ai/integrations/raycast/cohere-transcribe)
- [VS Code](https://spiral.ai/integrations/vscode/cohere-transcribe)
- [Windsurf](https://spiral.ai/integrations/windsurf/cohere-transcribe)
- [Zapier](https://spiral.ai/integrations/zapier/cohere-transcribe)

## More Factory Droid guides

- [Chatterbox TTS](https://spiral.ai/integrations/factory-droid/chatterbox-tts)
- [DeepFilterNet 3](https://spiral.ai/integrations/factory-droid/deepfilternet-3)
- [Chatterbox Speech to Speech](https://spiral.ai/integrations/factory-droid/chatterbox-speech-to-speech)
- [Demucs](https://spiral.ai/integrations/factory-droid/demucs)
- [Chatterbox Multilingual](https://spiral.ai/integrations/factory-droid/chatterbox-multilingual)
- [Eleven Flash v2](https://spiral.ai/integrations/factory-droid/eleven-flash-v2)

## Questions about Cohere Transcribe in Factory Droid

**Can Factory Droid use Cohere Transcribe?**

Yes. Add Spiral to Factory Droid as an MCP server, and Factory Droid can find and run Cohere Transcribe like any built-in tool.

**How much does Cohere Transcribe cost in Factory Droid?**

Up to $0.005382 per run at the default inputs for 60 seconds of input. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Factory Droid plan is separate.

**Do I need a fal account to use Cohere Transcribe in Factory Droid?**

No. Spiral runs Cohere Transcribe on its own fal account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Cohere Transcribe in Factory Droid?**

About two minutes: add the server, then sign in, then check it.

**Does Factory Droid wait while Cohere Transcribe runs?**

Cohere Transcribe runs in the background. Factory Droid starts the run, checks on it and gives you the json when it's ready.

**Which languages does Cohere Transcribe support?**

It supports English, French, German, Italian, Spanish, Portuguese, Greek, Dutch, Polish, Mandarin Chinese, Japanese, Korean, Vietnamese, and Arabic. Choose the language actually spoken in the recording.

**Can Cohere Transcribe detect languages or handle a bilingual conversation?**

It does not automatically detect the language, and results are inconsistent when a recording switches between languages. It works best with one specified language; where practical, separate passages by language before transcription.

**Does Cohere Transcribe include timestamps or speaker names?**

No. It returns a transcript without timestamps or speaker identification, so timed captions and speaker-attributed meeting notes need an additional processing step.

**Why might words appear during silence?**

Cohere warns that non-speech sounds and quiet background noise can turn into invented text. Apply a noise gate or voice activity detection before transcription, and check suspicious passages against the recording.

**Can Cohere Transcribe handle long recordings?**

Cohere's reference processor splits long audio into chunks and joins the resulting transcripts. That supports long-form transcription, but does not establish a universal maximum recording length for every implementation.

