---
title: Sound Effects API: Text to SFX at $0.001/s | ModelsLab
description: Sound effects API on Stable Audio Open 1.0: $0.001 per second, 3–15 s clips, mp3/wav/flac, about 13 s per effect. curl and Python examples.
url: https://modelslab.com/text-to-sfx
canonical: https://modelslab.com/text-to-sfx
type: website
component: Seo/TextToSFX
generated_at: 2026-11-09T02:46:18.986668Z
---

AudioGen

Sound Effects API — Text to SFX
---

Last updated Oct 7, 2026 · By ModelsLab Engineering

The ModelsLab sound effects API turns a text prompt into a sound effect. It runs Stable Audio Open 1.0 on our own GPUs and costs $0.001 per second of audio ($0.0047 minimum), so a 10-second effect is $0.01. Clips are 3 to 15 seconds in mp3, wav or flac, and take about 13 seconds to generate. POST your prompt to /api/v6/voice/sfx.

[Get API key](https://modelslab.com/register) [See pricing](https://modelslab.com/pricing)

- $0.001 per second of audio
- Stable Audio Open 1.0
- No per-effect charge on the $149 plan

What the Sound Effects API Does
---

### Text In, Sound Effect Out

Send a plain-English description — “glass bottle shattering on a concrete floor” — and the API returns an audio file of that sound. The model is Stable Audio Open 1.0 from Stability AI, run on ModelsLab’s own GPUs with 80 diffusion steps at 44.1 kHz stereo. You choose the length (3 to 15 seconds) and the format (mp3, wav or flac).

Billing is per second of audio: $0.001 a second with a $0.0047 minimum per effect, drawn from your plan. On the $149/month Open Source Unlimited plan there is no per-effect charge at all, which makes it a flat price for teams that generate sound libraries in bulk.

ElevenLabs Sound Effects is on the same API key too, through `/api/v7/voice/sound-generation` with model\_id `eleven_sound_effect` at $0.06 per generation. Request errors logged per endpoint are on the [API status page](/status).

- Stable Audio Open 1.0 over plain HTTP — no SDK to install
- $0.001 per second, $0.0047 minimum; a 10-second effect is $0.01
- 3 to 15 second clips in mp3, wav or flac, 44.1 kHz stereo
- About 13 seconds of GPU time per effect
- File URL in the response, fetch by id, or webhook delivery

Hear the Model
---

Generated on this endpoint on October 6, 2026 with duration 5, mp3 at 192k. Unedited: what you hear is what the API returned.

### Door creak

Prompt: “Heavy wooden door creaking open slowly in a stone hallway”

[Door creak sample](/audios/sfx/stable-audio-open/door-creak.mp3)[Download the mp3](/audios/sfx/stable-audio-open/door-creak.mp3)

### Thunderstorm

Prompt: “Thunderstorm with heavy rain and a close lightning strike”

[Thunderstorm sample](/audios/sfx/stable-audio-open/thunderstorm.mp3)[Download the mp3](/audios/sfx/stable-audio-open/thunderstorm.mp3)

### Laser shots

Prompt: “Sci-fi laser gun firing three quick shots”

[Laser shots sample](/audios/sfx/stable-audio-open/laser-shots.mp3)[Download the mp3](/audios/sfx/stable-audio-open/laser-shots.mp3)

### Glass shatter

Prompt: “Glass bottle shattering on a concrete floor”

[Glass shatter sample](/audios/sfx/stable-audio-open/glass-shatter.mp3)[Download the mp3](/audios/sfx/stable-audio-open/glass-shatter.mp3)

Sound effects API quick start
---

One POST with a prompt. cURL, Python and JavaScript.

### cURL — one sound effect

bash

```
<code>1curl -X POST 'https://modelslab.com/api/v6/voice/sfx' \
2  -H 'Content-Type: application/json' \
3  -d '{
4    "key": "YOUR_MODELSLAB_API_KEY",
5    "prompt": "Heavy wooden door creaking open slowly in a stone hallway",
6    "duration": 5,
7    "output_format": "mp3"
8  }'</code>
```

### Python — generate, then fetch if still rendering

Python

```
<code>1import time
2import requests
3

4KEY = "YOUR_MODELSLAB_API_KEY"
5

6result = requests.post(
7    "https://modelslab.com/api/v6/voice/sfx",
8    json={
9        "key": KEY,
10        "prompt": "Glass bottle shattering on a concrete floor",
11        "duration": 5,
12        "output_format": "wav",
13    },
14    timeout=60,
15).json()
16

17# With a clear GPU queue the file comes back in this response. Otherwise the
18# status is "processing": keep the fetch URL and poll it until the status changes.
19fetch_url = result.get("fetch_result")
20for _ in range(60):
21    if result.get("status") != "processing":
22        break
23    time.sleep(10)
24    result = requests.post(fetch_url, json={"key": KEY}, timeout=60).json()
25

26if result.get("status") == "success":
27    print(result["output"][0])
28else:
29    print("Not ready:", result.get("status"), result.get("message"))</code>
```

### JavaScript — with a webhook

JavaScript

```
<code>1const response = await fetch('https://modelslab.com/api/v6/voice/sfx', {
2  method: 'POST',
3  headers: { 'Content-Type': 'application/json' },
4  body: JSON.stringify({
5    key: 'YOUR_MODELSLAB_API_KEY',
6    prompt: 'Sci-fi laser gun firing three quick shots',
7    duration: 3,
8    // The finished file is POSTed here as well as returned or fetched.
9    webhook: 'https://your-app.example.com/hooks/sfx',
10    track_id: 'level-4-laser',
11  }),
12});
13

14console.log(await response.json());</code>
```

Specs and Limits
---

Everything an integration needs to know before the first call.

| Spec | ModelsLab sound effects API |
|---|---|
| Endpoint | POST /api/v6/voice/sfx |
| Model | Stable Audio Open 1.0 (Stability AI), 80 steps |
| Price | $0.001 per second of audio, $0.0047 minimum per effect |
| 10-second effect | $0.01 (no per-effect charge on the $149 plan) |
| Clip length | 3 to 15 seconds, whole seconds, default 8 |
| Formats | mp3 (default), wav, flac |
| Bitrate | 128k, 192k or 320k (default 320k) |
| Audio | 44.1 kHz, stereo |
| Typical speed | About 13 s per effect (12.9–13.3 s measured) |
| Input | Text prompt only |
| Delivery | File URL in the response, fetch by id, or webhook |

Read from the running endpoint on October 6, 2026. Speed is GPU time per effect from the worker logs; add network time and any queue wait.

Request and Response
---

The full contract of POST /api/v6/voice/sfx.

| Field | Type | Meaning |
|---|---|---|
| key | string (required) | Your ModelsLab API key |
| prompt | string (required) | The sound to generate, in plain English |
| duration | integer (optional) | Seconds of audio, 3 to 15; default 8 |
| output\_format | string (optional) | mp3, wav or flac; default mp3 |
| bitrate | string (optional) | 128k, 192k or 320k; default 320k |
| webhook | URL (optional) | Where to POST the result when it is ready |
| track\_id | string (optional) | Your own reference, echoed back |
| output | response | Array with the URL of the audio file |
| id, fetch\_result | response | Returned with status processing; POST the fetch URL to get the file |

The key goes in the JSON body. A response with status success carries the file URL in output. A response with status processing carries id, fetch\_result and future\_links; POST to fetch\_result with your key, or wait for the webhook.

What Sound Effects Cost
---

One ModelsLab subscription covers sound effects, speech, image and video.

| Plan | Price / month | Sound effects | Parallel requests | Also covers |
|---|---|---|---|---|
| Basic | $21 | $0.001 per second from included usage | 5 | Image, video, speech and LLM APIs |
| Standard | $47 | $0.001 per second from included usage | 10 | Image, video, speech and LLM APIs |
| Open Source Unlimited | $149 | No per-effect charge on Stable Audio Open | 15 | Every open-weight model ModelsLab self-hosts |

Plan prices from the live plans table. On Basic and Standard, each effect draws $0.001 per second of audio, with a $0.0047 minimum, from the plan's included usage.

Sound Effects API Providers Compared
---

Fourteen text-to-sound-effects offers, sorted by what one 10-second effect costs.

| Provider | Model | Listed price | 10-second effect | Max length |
|---|---|---|---|---|
| Pika (dev.pika.art) | Pika SFX | $0.0002 / second + $10/month membership | $0.002 | 20 s |
| Replicate | MMAudio (zsxkib/mmaudio) | ≈ $0.0049 / run (L40S GPU time) | ≈ $0.0049 | Not published |
| ModelsLab | Stable Audio Open 1.0 | $0.001 / second, $0.0047 minimum; none on the $149 plan | $0.01 | 15 s |
| fal | MMAudio V2 | $0.001 / second | $0.01 | 30 s |
| fal | Sonilo v1.1 | $0.0018 / second | $0.018 | 180 s |
| ElevenLabs | Sound Effects v2 (eleven\_text\_to\_sound\_v2) | $0.12 / minute of audio | $0.02 | 30 s |
| fal | ElevenLabs Sound Effects V2 | $0.002 / second | $0.02 | 22 s |
| fal | Stable Audio 3 Small SFX | $0.0206 / audio | $0.0206 | 120 s |
| WaveSpeed | Kling text-to-audio | $0.035 / run | $0.035 (max 10 s) | 10 s |
| Segmind | ElevenLabs sound generation | $0.05 / generation | $0.05 | Not published |
| ModelsLab | ElevenLabs Sound Effects (eleven\_sound\_effect) | $0.06 / generation | $0.06 | Not published |
| WaveSpeed | Mirelo SFX 1.6 | $0.01 / second | $0.10 | 60 s |
| Stability AI | Stable Audio 2.5 | 20 credits ($0.20) / generation | $0.20 | 190 s |
| fal | Stable Audio 2.5 | $0.20 / audio | $0.20 | 190 s |

Every price read on the provider's own page on October 6, 2026, in the unit the provider publishes. ElevenLabs lists Sound Effects at $0.12 per minute on its API pricing page, while its FAQ says sound effects are billed per generation and its docs say 40 credits per second when a duration is set. Replicate bills GPU time, so its cost per run varies. Pika's $10/month membership is charged on top of usage.

### Which sound effects API should you use?

Per second, Pika SFX is the cheapest offer we found at $0.0002, but it adds a $10/month membership. MMAudio V2 on fal and this endpoint both cost $0.001 a second; ElevenLabs costs $0.12 a minute, which is $0.002 a second. At volume the arithmetic changes: above about 14,900 ten-second effects a month, the $149 Open Source Unlimited plan costs less than any $0.001-per-second API.

Price is not the whole answer. ElevenLabs Sound Effects v2 allows 30-second clips and loops. MMAudio can generate audio from a video as well as from text. Stable Audio 2.5 and Sonilo run past two minutes. This endpoint is text only and stops at 15 seconds — use it for libraries of short effects, game one-shots and app sounds, and pick one of the others when length, video sync or the last bit of realism matters.

Sound effects price sources
---

Every competitor price and limit on this page, with the date we read it.

- [ElevenLabs API pricing](https://elevenlabs.io/pricing/api) — Sound Effects $0.12 per minute; FAQ "billed per generation", read 2026-10-06
- [ElevenLabs sound effects docs](https://elevenlabs.io/docs/overview/capabilities/sound-effects) — 40 credits per second with a duration, formats, read 2026-10-06
- [ElevenLabs sound effects API reference](https://elevenlabs.io/docs/api-reference/text-to-sound-effects/convert) — duration 0.5–30 s, eleven\_text\_to\_sound\_v2, read 2026-10-06
- [Pika SFX pricing](https://dev.pika.art/models/pika/pika-audio/pika-sfx/pricing) — $0.0002 per second, read 2026-10-06
- [Pika API pricing](https://dev.pika.art/pricing) — $10/month membership fee, read 2026-10-06
- [fal MMAudio V2 text-to-audio](https://fal.ai/models/fal-ai/mmaudio-v2/text-to-audio) — $0.001 per second, read 2026-10-06
- [fal ElevenLabs Sound Effects V2](https://fal.ai/models/fal-ai/elevenlabs/sound-effects/v2) — $0.002 per second, read 2026-10-06
- [fal Sonilo v1.1 text-to-sound-effects](https://fal.ai/models/sonilo/v1.1/text-to-sound-effects) — $0.0018 per second, read 2026-10-06
- [fal Stable Audio 3 Small SFX](https://fal.ai/models/fal-ai/stable-audio-3/small/sfx/text-to-audio) — $0.0206 per audio, read 2026-10-06
- [fal Stable Audio 2.5](https://fal.ai/models/fal-ai/stable-audio-25/text-to-audio) — $0.20 per audio, read 2026-10-06
- [WaveSpeed Mirelo SFX 1.6](https://wavespeed.ai/models/mirelo-ai/sfx-1.6/text-to-audio) — $0.01 per generated second, read 2026-10-06
- [WaveSpeed Kling text-to-audio](https://wavespeed.ai/models/kwaivgi/kling-text-to-audio) — $0.035 per run, read 2026-10-06
- [Stability AI pricing](https://platform.stability.ai/pricing) — Stable Audio 2.5 = 20 credits, 1 credit = $0.01, read 2026-10-06
- [Replicate MMAudio](https://replicate.com/zsxkib/mmaudio) — about $0.0049 per run on L40S, read 2026-10-06
- [Segmind sound generation pricing](https://segmind.com/models/sound-generation/pricing) — $0.05 per generation, read 2026-10-06
- [Stable Audio Open 1.0 model card](https://huggingface.co/stabilityai/stable-audio-open-1.0) — Stability AI Community License

Generate a Sound Effect in 3 Steps
---

From API key to an audio file.

STEP 01

STEP 01

### Step 1: Get Your API Key

Create a ModelsLab account, subscribe to a plan from $21/month, and copy your key from the dashboard.

STEP 02

STEP 02

### Step 2: POST a Prompt

Send key and prompt to /api/v6/voice/sfx. Add duration (3 to 15 seconds) and output\_format (mp3, wav or flac) when the defaults do not fit.

STEP 03

STEP 03

### Step 3: Download the Audio

With a clear GPU queue, the response carries the file URL in output. Otherwise POST to fetch\_result with your key until the status changes, or pass a webhook and receive the file when it is ready.

[Get API Key ](https://modelslab.com/register)

Sound effects API guides
---

[### Sound effects API pricing

Price per second and per 10-second effect across 14 offers.](/sound-effects-api-pricing) [### Best sound effects API 2026

ElevenLabs, Pika, MMAudio, Stable Audio and ModelsLab compared.](/best-sound-effects-api-2026) [### ElevenLabs sound effects alternative

What switching costs, what you gain and what you lose.](/elevenlabs-sound-effects-api-alternative) [### API status

Request errors logged per endpoint, last seven days.](/status)

Why Generate Sound Effects Here
---

Key advantages that set us apart

Stable Audio Open 1.0 on our own GPUs

$0.001 per second of audio, $0.0047 minimum

No per-effect charge on the $149 plan

3 to 15 second clips in mp3, wav or flac

ElevenLabs Sound Effects on the same key

One key for sound, speech, image, video and LLMs

Plans from $21/month

Our Popular Use Cases

What teams build with the sound effects API:

Game Audio LibrariesVideo and ShortsApp and Product SoundsPodcasts and Audio StoriesPrototyping Sound DesignGenerative Media Pipelines

Generate footsteps, impacts, pickups and UI sounds in bulk, then keep the best variation of each.

![Game Audio Libraries](https://imagedelivery.net/PP4qZJxMlvGLHJQBm3ErNg/0fbacb1a-6e34-4254-0a9d-5e75178cf200/768)

Your Data is Secure: GDPR Compliant AI Services
---

![ModelsLab GDPR Compliance Certification Badge](https://imagedelivery.net/PP4qZJxMlvGLHJQBm3ErNg/28133112-07fe-4c1c-44eb-36948d51ae00/768)

Pricing That's Perfect
---

Choose plan as per your needs, cancel anytime.

100% refund policy on monthly & yearly plans — cancel anytime

MonthlyYearly 2 Months Free

 [Contact Sales](https://calendly.com/support-lael/30min)

Best Value

### Open Source Unlimited

Mission-Critical

$149 /month

Get API Key100% refund policy · cancel anytime

Unlimited Open Source Models

100% refund policy

24x7 Support

15 parallel generations

Access to all APIs

Unlimited generations on all open-source models

For mission critical workloads

Add Team Members

Priority GPU Clusters

Most Popular

### Standard

Production

$47 /month

Get API Key100% refund policy · cancel anytime

Moderate Traffic

100% refund policy

Priority Developer Support

10 concurrent API requests

For Production workloads

API access to all models

Prototype

### Basic

Prototype

$21 /month

Get API Key100% refund policy · cancel anytime

Moderate Traffic

100% refund policy

Developer Support via Discord/Email

5 concurrent API requests

API access to all models

Shared GPU

TestimonialsTrusted by 450+ teams on dedicated GPUs
---

See how developers and engineering teams scale generative pipelines with ModelsLab.

Previous slideNext slide

ModelsLab's Voice Cloning and Audio APIs revolutionized how we approach character development in our games. It's like having a full studio of voice actors running on dedicated infrastructure.

![Alex Rivera](/images/testimonials/alex-rivera.svg)

#### Alex Rivera

Lead Game Developer at Voxel Studios

The LLM and image generation APIs made multi-model pipelines trivial. One unified API key replaced four separate vendor contracts and cut our queue latency in half.

![John Hernandez](/images/testimonials/john-hernandez.svg)

#### John Hernandez

AI Systems Architect at Synthetix AI

The ease of creating lifelike voiceovers for our e-learning courses has dramatically increased engagement. 99.9% uptime and instant API response have made it essential to our stack.

![Priya Singh](/images/testimonials/priya-singh.svg)

#### Priya Singh

Instructional Designer at LearnStream Media

Dedicated GPU clusters gave our production pipelines the throughput we needed without unpredictable rate limits or cold starts. Excellent developer experience.

![Michael Chen](/images/testimonials/michael-chen.svg)

#### Michael Chen

Digital Marketing Manager at Nexus Studio

Related Sound Effects APIs
---

- [Sound effects API pricing](/sound-effects-api-pricing)Price per second across 14 offers.
- [Best sound effects API](/best-sound-effects-api-2026)ElevenLabs, Pika, MMAudio and ModelsLab compared.
- [ElevenLabs sound effects alternative](/elevenlabs-sound-effects-api-alternative)Half the per-second rate, and what you give up.
- [Cheapest AI API](/cheapest-ai-api)Image, video and speech prices side by side.
- [Unlimited AI API](/unlimited)Every self-hosted open-source model for $149/mo.

Get expert support

We're Here to Help.
---

Want to know more? You can email us anytime at <support@modelslab.com>

Chat with support[View Docs](https://docs.modelslab.com)


POST https://modelslab.com/api/v6/voice/sfx with your `key` and a `prompt`. Optional fields: `duration` (a whole number of seconds from 3 to 15, default 8), `output_format` (mp3, wav or flac, default mp3), `bitrate` (128k, 192k or 320k, default 320k), `webhook` and `track_id`. The response carries the audio URL in `output`, or an `id` to fetch from /api/v6/voice/fetch/{id} while the clip is still rendering.


Stable Audio Open 1.0 from Stability AI, running on ModelsLab’s own GPUs with 80 diffusion steps and 44.1 kHz stereo output. ElevenLabs Sound Effects is also available on the same API key through POST /api/v7/voice/sound-generation with model_id eleven_sound_effect.


$0.001 per second of audio, with a minimum of $0.0047 per effect, drawn from the usage included in your plan: a 10-second effect costs $0.01. Plans start at $21/month. On the $149/month Open Source Unlimited plan, Stable Audio Open effects carry no per-effect charge. ElevenLabs Sound Effects (eleven_sound_effect) costs $0.06 per generation.


About 13 seconds of GPU time per effect, whatever its length: the worker logged 12.9 to 13.3 seconds across 20 effects on October 6, 2026. When the GPU queue is clear, the API call returns the finished file. When it is not, the response has status processing, an id and the future file URL: POST to the fetch URL with your key until status is success, failed or error, or pass a webhook.


3 to 15 seconds per request, in whole seconds. For longer ambience, generate overlapping clips and crossfade them, or use a model built for long audio such as Stable Audio 2.5, which runs up to 190 seconds on Stability AI’s API.


ModelsLab charges no royalties or licence fees on the audio you generate. Stable Audio Open 1.0 is released by Stability AI under the Stability AI Community License; read it for the terms that apply to your own use.


modelslab.com/status lists every API endpoint, including /api/v6/voice/sfx, with the request errors logged over the last seven days. It does not yet count effects that time out in the GPU queue, so check your own responses for status failed or error as well.


No. The endpoint turns text into audio. For foley synced to footage, use a video-to-audio model such as MMAudio, Mirelo SFX or Hunyuan Video Foley.

Explore Our Other Solutions
---

Unlock your creative potential and scale your business with ModelsLab's comprehensive suite of AI-powered solutions.

[Imagen

### AI Image Generation & Tools

Generate, edit, upscale, and transform images with state-of-the-art AI models.

Explore Imagen](https://modelslab.com/imagen) [Video Fusion

### AI Video Generation & Tools

Create, edit, and enhance videos with AI-powered generation and transformation tools.

Explore Video Fusion](https://modelslab.com/video-generation) [Chat

### Engage Seamlessly with LLM

Access powerful language models for chatbots, content generation, and AI assistants.

Explore Chat](https://modelslab.com/chatbot-api) [3D Verse

### Create Stunning 3D Models

Transform images and text into 3D models with advanced AI-powered generation.

Explore 3D Verse](https://modelslab.com/text-to-3d)

Plugins

Explore Plugins for Pro
---

Our plugins are designed to work with the most popular content creation software.

[Explore Plugins](https://modelslab.com/pro#plugins) [Learn More](https://modelslab.com/pro)

API

Build Apps with ModelsLab

ML

 API
---

Use our API to build apps, generate AI art, create videos, and produce audio with ease.

[API Documentation](https://docs.modelslab.com) [Playground](https://modelslab.com/models)

## Frequently Asked Questions

### Which endpoint generates sound effects?
POST https://modelslab.com/api/v6/voice/sfx with your `key` and a `prompt`. Optional fields: `duration` (a whole number of seconds from 3 to 15, default 8), `output_format` (mp3, wav or flac, default mp3), `bitrate` (128k, 192k or 320k, default 320k), `webhook` and `track_id`. The response carries the audio URL in `output`, or an `id` to fetch from /api/v6/voice/fetch/{id} while the clip is still rendering.

### Which model generates the sound effects?
Stable Audio Open 1.0 from Stability AI, running on ModelsLab’s own GPUs with 80 diffusion steps and 44.1 kHz stereo output. ElevenLabs Sound Effects is also available on the same API key through POST /api/v7/voice/sound-generation with model_id eleven_sound_effect.

### How much does the sound effects API cost?
$0.001 per second of audio, with a minimum of $0.0047 per effect, drawn from the usage included in your plan: a 10-second effect costs $0.01. Plans start at $21/month. On the $149/month Open Source Unlimited plan, Stable Audio Open effects carry no per-effect charge. ElevenLabs Sound Effects (eleven_sound_effect) costs $0.06 per generation.

### How long does one sound effect take to generate?
About 13 seconds of GPU time per effect, whatever its length: the worker logged 12.9 to 13.3 seconds across 20 effects on October 6, 2026. When the GPU queue is clear, the API call returns the finished file. When it is not, the response has status processing, an id and the future file URL: POST to the fetch URL with your key until status is success, failed or error, or pass a webhook.

### How long can a generated sound effect be?
3 to 15 seconds per request, in whole seconds. For longer ambience, generate overlapping clips and crossfade them, or use a model built for long audio such as Stable Audio 2.5, which runs up to 190 seconds on Stability AI’s API.

### Can I use the generated sound effects commercially?
ModelsLab charges no royalties or licence fees on the audio you generate. Stable Audio Open 1.0 is released by Stability AI under the Stability AI Community License; read it for the terms that apply to your own use.

### Is there a status page for the sound effects API?
modelslab.com/status lists every API endpoint, including /api/v6/voice/sfx, with the request errors logged over the last seven days. It does not yet count effects that time out in the GPU queue, so check your own responses for status failed or error as well.

### Can it generate sound for an existing video?
No. The endpoint turns text into audio. For foley synced to footage, use a video-to-audio model such as MMAudio, Mirelo SFX or Hunyuan Video Foley.


---

*This markdown version is optimized for AI agents and LLMs.*

**Links:**
- [Website](https://modelslab.com)
- [API Documentation](https://docs.modelslab.com)
- [Blog](https://modelslab.com/blog)

---
*Generated by ModelsLab - 2026-11-09*