[Integrations](https://spiral.ai/integrations) › [opencode](https://spiral.ai/integrations/opencode) › [SAM Audio Visual Separation](https://spiral.ai/integrations/opencode/sam-audio-visual-separation)

# How to Use SAM Audio Visual Separation in opencode

SAM Audio Visual Separation extracts the sound of a selected person, instrument, or object from video. Visual masks identify the source, and a short sound description can help specify what to isolate.

|  |  |
| --- | --- |
| Model | SAM Audio Visual Separation |
| Available through | fal |
| Category | Audio |
| Price | $0.13 per 30 seconds of audio for its two tracks, at least 30 seconds. Each extra reranking candidate: $0.065 per 30 seconds. |
| Returns | audio |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with SAM Audio Visual Separation in opencode

- Use SAM Audio Visual Separation on the interview footage and matching speaker masks in this project. Isolate the interviewee with the description 'woman speaking' and save an original-versus-extraction review pair.
- Use SAM Audio Visual Separation on the performance video and guitar mask in the project. Describe the target as 'guitar playing' and prepare the extracted part for a practice review.
- Use SAM Audio Visual Separation on the wildlife clips with their matching animal masks. Extract 'dog barking' from the relevant clips and organize the results beside the originals.
- Use SAM Audio Visual Separation on the vehicle footage and matching masks in this repository. Extract 'engine revving' and create a listening comparison for the sound editor.
- Use SAM Audio Visual Separation to compare one candidate with four reranked candidates on the project's cooking clip and pan mask. Use 'pan sizzling' as the description and prepare both results for review.

## How to set up SAM Audio Visual Separation in opencode

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

### Use Spiral's MCP server

1. **Add the server**: Leave out `--global` to add Spiral to the current project only. No header and no client ID or secret: opencode registers itself with Spiral.

```bash
opencode mcp add spiral --global --url https://api.spiral.ai/mcp
```

2. **Sign in**: Your browser opens. Sign in to Spiral, pick the organization and click Authorize, then return to the terminal. No API key. Inside opencode, `/mcps` starts the same sign-in.

```bash
opencode mcp auth spiral
```

3. **Check it**: Spiral shows as connected.

```bash
opencode mcp list
```

### Use the Spiral skill

1. **Send opencode 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 SAM Audio Visual Separation**: 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/sam-audio/visual-separate",
    "input": {
      "video_url": "https://v3b.fal.media/files/b/0a8850d1/gff7zKI-6XwIbBBip4946_office.mp4"
    }
  }'
```

Then paste one of the prompts above: opencode finds SAM Audio Visual Separation on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/opencode

## SAM Audio Visual Separation tools and prices

- [SAM Audio · Video-Prompted Sound Separation](https://spiral.ai/tools/fal/fal-ai/sam-audio/visual-separate) (fal): $0.13 per 30 seconds of audio for its two tracks, at least 30 seconds. Each extra reranking candidate: $0.065 per 30 seconds.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| mask_video_url |  | — | URL of the mask video (binary mask indicating target object). Black=target, White=background. |
| prompt |  |  | Text prompt to assist with separation. Use natural language to describe the target sound. |
| reranking_candidates | 1 to 4 | 1 | Number of candidates to generate and rank. Higher improves quality but increases latency and cost. |
| acceleration | fast, balanced, quality | balanced | The acceleration level to use. |
| max_chunk_duration | 10 to 60 | 60 | Maximum audio duration (in seconds) to process in a single pass. Longer audio will be chunked with overlap and blended. |
| chunk_overlap | 0 to 30 | 5 | Overlap duration (in seconds) between chunks for crossfade blending. |

## What can you do with SAM Audio Visual Separation in opencode?

Turn project footage into isolated dialogue, instrument parts, and sound effects for editing. A coding agent can organize extraction jobs around selected sources and prepare comparisons for listening review.

### Interview dialogue

Extract a selected interviewee's speech from video for an editor to review before assembling dialogue.

### Instrument practice

Isolate a visible guitarist or pianist in a performance video to study the recorded part.

### Wildlife sound clips

Extract a selected animal's audible call from field footage for a nature edit.

### Vehicle sound effects

Separate a visible vehicle's engine sound from a recorded scene for sound design review.

### Cooking demonstration audio

Extract the sound of a selected pan sizzling or knife chopping from cooking footage.

### Documentary sound editing

Pull a visible sound source, such as a ringing bell or running machine, from location footage for an editor to audition.

## Overview

- [fal](https://spiral.ai/providers/fal)
- [All opencode guides](https://spiral.ai/integrations/opencode)

## SAM Audio Visual Separation in other agents

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

## More opencode guides

- [SAM Audio](https://spiral.ai/integrations/opencode/sam-audio)
- [Scribe v2](https://spiral.ai/integrations/opencode/scribe-v2)
- [Qwen3-TTS 1.7B Voice Design](https://spiral.ai/integrations/opencode/qwen3-tts-1-7b-voice-design)
- [Scribe v3](https://spiral.ai/integrations/opencode/scribe-v3)
- [Qwen3-TTS 1.7B](https://spiral.ai/integrations/opencode/qwen3-tts-1-7b)
- [Seed Audio 1.0](https://spiral.ai/integrations/opencode/seed-audio-1-0)

## Questions about SAM Audio Visual Separation in opencode

**Can opencode use SAM Audio Visual Separation?**

Yes. Add Spiral to opencode as an MCP server, and opencode can find and run SAM Audio Visual Separation like any built-in tool.

**How much does SAM Audio Visual Separation cost in opencode?**

$0.13 per 30 seconds of audio for its two tracks, at least 30 seconds. Each extra reranking candidate: $0.065 per 30 seconds. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your opencode plan is separate.

**Do I need a fal account to use SAM Audio Visual Separation in opencode?**

No. Spiral runs SAM Audio Visual Separation on its own fal account. You sign in to Spiral once and never paste a key.

**How long does it take to set up SAM Audio Visual Separation in opencode?**

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

**Does opencode wait while SAM Audio Visual Separation runs?**

SAM Audio Visual Separation runs in the background. opencode starts the run, checks on it and gives you the audio when it's ready.

**What sounds can SAM Audio isolate?**

SAM Audio separates speech, music, individual instruments, and everyday sounds. Visual separation lets you identify the person or object producing the sound in a video.

**How should I describe the sound I want?**

Use a short, lowercase phrase such as 'woman speaking', 'guitar playing', or 'dog barking'. Describe the sound to keep, with the visual selection identifying its source.

**Can SAM Audio isolate one singer from a chorus?**

Highly similar sounds remain difficult to separate, including one singer in a chorus or one instrument in an orchestra. Listen to the extraction before using it in a final edit.

**Is visual selection enough for a clean extraction?**

A visual selection identifies the intended source, but Meta reports weaker results from visual prompting than text prompting. Try adding a short description of the target sound and compare the results.

**What should I check in a long recording?**

Listen across processing boundaries as well as within each section. Meta found that long recordings could lose quality in a single pass, while independently processed chunks could produce audible discontinuities.

