# Seed ASR 2.0 — Atlas Cloud API

> BytePlus Seed Speech recording-file recognition (ASR 2.0): transcribe an audio file to text with punctuation, ITN, sentence segmentation, speaker diarization, and 51-language support.

This is the machine-readable API reference for **Seed ASR 2.0** on Atlas Cloud,
a unified API platform for 400+ AI models across text, image, video, audio and 3D.

- **Model ID**: `bytedance/seed-asr-2.0`
- **Built by**: ByteDance
- **Modality**: Audio
- **Model page**: https://www.atlascloud.ai/models/bytedance/seed-asr-2.0
- **API key**: https://www.atlascloud.ai/console/api-keys
- **Docs**: https://www.atlascloud.ai/docs

## Pricing on Atlas Cloud

- $0.002 per minute of audio
- Pay-as-you-go. No minimum spend, no subscription required.

> **These are the authoritative Atlas Cloud rates for this model.** Any price that
> appears in the vendor description further down refers to a different platform or
> a different model variant and does not apply here.

## Use this model from an AI agent

Atlas Cloud ships three first-party integration surfaces. All three authenticate
with the same API key via the `ATLASCLOUD_API_KEY` environment variable.

### MCP server

The official MCP server (`atlascloud-mcp`) exposes this model to any
MCP-compatible host — Claude Code, OpenAI Codex, Cursor, Gemini CLI, Goose,
Claude Desktop. One-line install:

```bash
# Claude Code
claude mcp add atlascloud -- npx -y atlascloud-mcp

# OpenAI Codex CLI
codex mcp add atlascloud -- npx -y atlascloud-mcp

# Gemini CLI
gemini mcp add atlascloud -- npx -y atlascloud-mcp

export ATLASCLOUD_API_KEY="your-api-key"
```

Then ask in plain English; the agent calls `atlas_generate_audio` with `model: "bytedance/seed-asr-2.0"`.
The server fetches each model's schema and validates parameters before submitting,
so invalid requests fail fast without spending credits.

MCP docs: https://www.atlascloud.ai/docs/mcp-server

### Agent Skills

`atlas-cloud-skills` is a portable skill package (API reference, code templates in
Python / Node.js / cURL, model IDs with pricing) for Claude Code, Cursor, Codex and
12+ other agents:

```bash
npx skills add AtlasCloudAI/atlas-cloud-skills
export ATLASCLOUD_API_KEY="your-api-key"
```

Skills docs: https://www.atlascloud.ai/docs/skills

### CLI

The `atlas` binary runs Atlas Cloud from a terminal or CI script. Async media jobs
are polled and downloaded automatically (use `--no-download` when a script only
needs the output URLs):

```bash
# Install (Homebrew, npm, or shell installer)
brew install AtlasCloudAI/tap/atlascloud
# npm install -g atlascloud-cli
# curl -fsSL https://raw.githubusercontent.com/AtlasCloudAI/cli/main/install.sh | sh

atlas auth login
atlas models get bytedance/seed-asr-2.0 --json
```

CLI docs: https://www.atlascloud.ai/docs/cli

## HTTP API reference

- **Submit endpoint (POST)**: `https://api.atlascloud.ai/api/v1/model/generateAudio` — start an async generation; returns a `prediction_id`
- **Poll endpoint (GET)**: `https://api.atlascloud.ai/api/v1/model/prediction/{prediction_id}` — poll this until the prediction finishes
- **Model ID**: `bytedance/seed-asr-2.0`


## API Information

This model can be used via our HTTP API or more conveniently via our client libraries.
See the input and output schema below, as well as the usage examples.


### Input Schema

The API accepts the following input parameters:

- **`model`** (`string`, _required_):
  model name
  - Default: `"bytedance/seed-asr-2.0"`

- **`audio_url`** (`string`, _required_):
  URL of the audio file to transcribe. Public URL or Base64. Formats: wav/mp3/ogg/raw. Max ~512MB.

- **`format`** (`string`, _optional_):
  Audio container format.
  - Default: `"mp3"`
  - Options: "mp3", "wav", "ogg", "raw"

- **`language`** (`string`, _optional_):
  Optional language code. Leave empty to auto-recognize Chinese/English/dialects. Or set a code, e.g. en-US, zh-CN, yue-CN, ja-JP, es-MX (51 languages supported).

- **`enable_itn`** (`boolean`, _optional_):
  Inverse text normalization (e.g. 'one hundred' -> '100').
  - Default: `true`

- **`enable_punc`** (`boolean`, _optional_):
  Add punctuation to the transcript.
  - Default: `false`

- **`enable_ddc`** (`boolean`, _optional_):
  Semantic smoothing (remove fillers/repeats).
  - Default: `false`

- **`enable_speaker_info`** (`boolean`, _optional_):
  Speaker diarization (<=10 speakers).
  - Default: `false`

- **`show_utterances`** (`boolean`, _optional_):
  Return per-sentence segments with timestamps.
  - Default: `false`

- **`context`** (`string`, _optional_):
  Inline hotwords, no table needed. Must be a JSON STRING of the form {"hotwords":[{"word":"YourTerm"}]}. WARNING: plain non-JSON text makes the task FAIL; unrecognized JSON keys are silently ignored.

- **`boosting_table_name`** (`string`, _optional_):
  Hotword table name to boost recognition of listed terms (e.g. product names). Tables are provisioned server-side by Atlas — contact us with your word list to get a table name. NOTE: a nonexistent table name is silently ignored (the task still succeeds).

- **`correct_table_name`** (`string`, _optional_):
  Correction table name: post-recognition forced replacement (wrong→right pairs). Provisioned server-side by Atlas, same as boosting_table_name. A nonexistent table name is silently ignored.



**Required Parameters Example**:

```json
{
  "model": "bytedance/seed-asr-2.0",
  "audio_url": ""
}
```


**Full Example**:

```json
{
  "model": "bytedance/seed-asr-2.0",
  "audio_url": "",
  "format": "mp3",
  "language": "",
  "enable_itn": true,
  "enable_punc": false,
  "enable_ddc": false,
  "enable_speaker_info": false,
  "show_utterances": false,
  "context": "",
  "boosting_table_name": "",
  "correct_table_name": ""
}
```


### Output Schema

The API returns the following output format:


- **`model`** (`string`, _optional_):
  model name
  - Default: `"bytedance/seed-asr-2.0"`

- **`created_at`** (`string`, _optional_):
  ISO timestamp of when the request was created.

- **`id`** (`string`, _optional_):
  Unique identifier for the prediction.

- **`outputs`** (`array[string]`, _optional_):
  Array with the recognized transcript text (outputs[0]).

- **`status`** (`string`, _optional_):
  Status: created, processing, completed, or failed.

- **`urls`** (`object`, _optional_):
  Object containing related API endpoints.

- **`stt_result`** (`object`, _optional_):
  Structured transcription result: text, duration (seconds), and words[] with per-word timestamps when show_utterances=true.
  - Properties:
    - **`text`** (`string`, _optional_):
      Full transcript.

    - **`duration`** (`number`, _optional_):
      Audio duration in seconds.

    - **`words`** (`array[object]`, _optional_):
      Word-level segments (show_utterances=true). Times in seconds.
      - Item properties:
        - **`text`** (`string`, _optional_):

        - **`start`** (`number`, _optional_):

        - **`end`** (`number`, _optional_):

        - **`type`** (`string`, _optional_):
          word | utterance

        - **`speaker_id`** (`string`, _optional_):
          Present when enable_speaker_info=true.





**Example Response**:

```json
{
  "model": "bytedance/seed-asr-2.0",
  "created_at": "",
  "id": "",
  "outputs": [
    ""
  ],
  "status": "",
  "urls": {},
  "stt_result": {
    "text": "",
    "duration": 0,
    "words": [
      {
        "text": "",
        "start": 0,
        "end": 0,
        "type": "",
        "speaker_id": ""
      }
    ]
  }
}
```


## Usage Examples

### cURL

```bash
# Step 1: Start generation (async)
curl -X POST "https://api.atlascloud.ai/api/v1/model/generateAudio" \
  -H "Authorization: Bearer $ATLASCLOUD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "bytedance/seed-asr-2.0",
  "audio_url": "",
  "format": "mp3",
  "language": "",
  "enable_itn": true,
  "enable_punc": false,
  "enable_ddc": false,
  "enable_speaker_info": false,
  "show_utterances": false,
  "context": "",
  "boosting_table_name": "",
  "correct_table_name": ""
}'

# Response will contain: {"code": 200, "data": {"id": "prediction_id", "status": "processing"}}

# Step 2: Poll for result (replace {prediction_id} with the id returned above)
curl -X GET "https://api.atlascloud.ai/api/v1/model/prediction/{prediction_id}" \
  -H "Authorization: Bearer $ATLASCLOUD_API_KEY"

# Keep polling until status is "completed", "succeeded" or "failed"
# When completed, outputs will contain the generated content URL(s)
```

## Additional Resources

### Documentation

- [Model Playground](https://www.atlascloud.ai/models/bytedance/seed-asr-2.0)

## About this model

_Vendor-supplied description. Any pricing or endpoint mentioned below refers to_
_other platforms — use the Atlas Cloud values above._

### Seed ASR 2.0 (Recording-File Recognition)

Transcribe an audio file to text with punctuation, ITN, sentence segmentation, speaker diarization, and 51-language support.

#### 🌟 Key Features

##### 🗣 High-Accuracy Transcription
State-of-the-art recognition tuned for real-world recordings.

##### 🌍 51 Languages
Auto-detects Chinese/English/dialects, or pin a language code.

##### 🎯 Rich Output
Punctuation, inverse text normalization, per-word timestamps, speaker labels.

##### 📻 Far-Field Optimized
Robust to meeting-room and call-center audio.

#### ⚙️ Parameters

| Parameter | Required | Description |
|---|---|---|
| `audio_url`* | ✅ | Audio file URL or Base64. Formats: wav/mp3/ogg/raw. |
| `format` | ⬜ | Audio format: `mp3` (default) / `wav` / `ogg` / `raw`. |
| `language` | ⬜ | Language code (e.g. `en-US`, `yue-CN`); empty = auto (CN/EN/dialects). 51 languages. |
| `enable_itn` | ⬜ | Inverse text normalization. Default `true`. |
| `enable_punc` | ⬜ | Add punctuation. Default `false`. |
| `enable_ddc` | ⬜ | Semantic smoothing. Default `false`. |
| `enable_speaker_info` | ⬜ | Speaker diarization. Default `false`. |
| `show_utterances` | ⬜ | Per-word timestamps in `stt_result.words`. Default `false`. |
| `context` | ⬜ | Inline hotwords — no table needed. JSON **string**: `{"hotwords":[{"word":"YourTerm"}]}`. See below. |
| `boosting_table_name` | ⬜ | Server-side hotword table name (provisioned by Atlas). |
| `correct_table_name` | ⬜ | Server-side correction table name (wrong→right replacement, provisioned by Atlas). |

#### 🔥 Hotwords & Correction

Two ways to improve recognition of product names, people, and domain terms:

**1. Inline via `context` (recommended — self-service, per request):**

```json
{
  "audio_url": "https://…/meeting.mp3",
  "context": "{\"hotwords\":[{\"word\":\"Zylophex\"},{\"word\":\"AtlasCloud\"}]}"
}
```

_(Description truncated. Full text on the model page.)_

---

Atlas Cloud — one API for 400+ AI models. Model page: https://www.atlascloud.ai/models/bytedance/seed-asr-2.0 · Docs: https://www.atlascloud.ai/docs