OmniHuman 2.0 API

bytedance/omnihuman-2.0/talking-avatar

OmniHuman 2.0 talking avatar turns one portrait and an audio file into a 720p clip of that person speaking or singing it. The clip is as long as your audio, from 4 to 300 seconds.

From $0.108/s

Run model View API

Send a request

Examples use the selected endpoint’s public parameter definitions. Replace the API key and example media, then send the request from your server.

SDKs & integrations

POSThttps://api.spicyapi.ai/api/v1/jobs/createTask

API authentication

Keep the API key in the server environment variable SPICY_API_KEY. Use Bearer authentication and Content-Type: application/json for JSON requests.

Authorization: Bearer $SPICY_API_KEY

Sample output

# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${SPICY_IDEMPOTENCY_KEY:-$(uuidgen)}"
curl -X POST https://api.spicyapi.ai/api/v1/jobs/createTask \
  -H "Authorization: Bearer $SPICY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $SPICY_IDEMPOTENCY_KEY" \
  -d '{
  "model": "bytedance/omnihuman-2.0/talking-avatar",
  "input": {
    "audio_url": "https://example.com/reference.mp3",
    "image_url": "https://example.com/reference.png",
    "prompt": "A slow aerial shot drifting over a misty pine forest at sunrise"
  }
}'

# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
  -H "Authorization: Bearer $SPICY_API_KEY"

Generated from this model’s input schema — the same one the API validates every request against.

Input parameters

Place these fields inside the request body’s input object. Names and types match the API. Do not assume a default when none is specified.

audio_urlstringRequired

Speech or singing, WAV or MP3, under 15 MB, 4 to 300 seconds. The clip is as long as this audio; audio outside that range is refused before anything is charged.

Default: Not specified
Constraints: format:"uri"
contentMediaType:["audio/mpeg","audio/wav"]

image_urlstringRequired

Portrait to animate. JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC or HEIF, under 30 MB, 300 to 6000 px on each side, width-to-height ratio between 0.4 and 2.5. We tested 16:9 portraits.

Default: Not specified
Constraints: format:"uri"
contentMediaType:["image/jpeg","image/png","image/webp","image/gif"]

promptstringOptional

Optional, up to 300 characters. Direct mood, gestures, where the eyes go and camera movement. Leave it empty and the model follows the audio on its own.

Default: Not specified
Constraints: maxLength:300

{
  "additionalProperties": false,
  "properties": {
    "audio_url": {
      "contentMediaType": [
        "audio/mpeg",
        "audio/wav"
      ],
      "description": "Speech or singing the portrait performs, WAV or MP3, under 15 MB, 4 to 300 seconds. The clip is as long as this audio.",
      "format": "uri",
      "type": "string",
      "x-ui": {
        "order": 88,
        "widget": "upload"
      }
    },
    "image_url": {
      "contentMediaType": [
        "image/jpeg",
        "image/png",
        "image/webp",
        "image/gif"
      ],
      "description": "Portrait to bring to life: JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC or HEIF, under 30 MB, 300 to 6000 px per side, width-to-height ratio 0.4 to 2.5.",
      "format": "uri",
      "type": "string",
      "x-ui": {
        "accept": [
          "image/jpeg",
          "image/png",
          "image/webp",
          "image/gif"
        ],
        "max_size_mb": 30,
        "order": 95,
        "widget": "upload"
      }
    },
    "prompt": {
      "description": "Optional direction for the performance: mood, gestures, where the eyes go, camera movement. Up to 300 characters.",
      "maxLength": 300,
      "type": "string",
      "x-ui": {
        "order": 100,
        "primary": true,
        "widget": "textarea"
      }
    }
  },
  "required": [
    "image_url",
    "audio_url"
  ],
  "type": "object",
  "x-pricing": {
    "billingMode": "duration_capped",
    "variantFields": []
  }
}

Track the task

HTTP 202 means the task was accepted, not completed. Save the returned taskId and query that task. Do not create another task while waiting for its result.

Accepted response

HTTP/1.1 202 Accepted
Location: /api/v1/jobs/recordInfo?taskId=job_01k3m8x9q2z4v7n5p6r8s0t1w3

{
  "code": 200,
  "msg": "success",
  "request_id": "req_example_create",
  "data": {
    "taskId": "job_01k3m8x9q2z4v7n5p6r8s0t1w3",
    "state": "queued",
    "estimatedCost": "0.750000"
  }
}

The accepted response includes the estimate so your app can show the same amount in task history and billing.

Read the job

curl --get "https://api.spicyapi.ai/api/v1/jobs/recordInfo" \
  --data-urlencode "taskId=$TASK_ID" \
  -H "Authorization: Bearer $SPICY_API_KEY"

Start polling at about 2 seconds and gradually increase the interval. Stop at a terminal state. After a connection failure, query the original taskId. Retry creation only with the original idempotency key and body.

Or let the callback find you

import { verifyWebhook } from "@spicyapi/sdk";

// express.raw keeps the bytes we signed; express.json never matches.
app.post("/hooks/spicy", express.raw({ type: "application/json" }), (req, res) => {
  const { payload } = verifyWebhook({
    rawBody: req.body,
    timestamp: req.header("X-Webhook-Timestamp"),
    signature: req.header("X-Webhook-Signature"),
    payloadVersion: Number(req.header("X-Webhook-Payload-Version")),
    secret: process.env.SPICY_WEBHOOK_SECRET,
  });
  res.status(200).end();
  void handleFinishedTask(payload);
});

Pass callBackUrl when you create the task and the finished record is POSTed to you — no polling loop at all. Every delivery carries X-Webhook-Timestamp, X-Webhook-Signature and X-Webhook-Payload-Version. The signature is a base64 HMAC-SHA256 over taskId.timestamp.sha256(body), so it must be checked against the raw bytes: parsing the JSON and re-serialising it changes them and the check will fail. Failed deliveries are retried, so the same taskId can arrive twice — make the handler idempotent. Answer 2xx first and do the work afterwards; a slow handler is treated as a failed delivery.

Read the result

Consume output only when state is succeeded. On failure, use errorCode, errorMessage and request_id to investigate. Example costs and states do not represent an actual bill.

Completed response example

{
  "code": 200,
  "msg": "success",
  "request_id": "req_example_record",
  "data": {
    "taskId": "job_01k3m8x9q2z4v7n5p6r8s0t1w3",
    "model": "bytedance/omnihuman-2.0/talking-avatar",
    "state": "succeeded",
    "output": {
      "assets": [
        {
          "key": "tasks/2026/08/28/job_01k3m8x9q2z4v7n5p6r8s0t1w3/3f7a1c9e5b2d48f0a6c1e3d5b7f9a1c3.mp4",
          "mime": "video/mp4",
          "url": "https://example.r2.cloudflarestorage.com/results/job_01k3m8x9q2z4v7n5p6r8s0t1w3/result.mp4?X-Amz-Signature=SIGNATURE_FROM_RESPONSE",
          "expiresAt": "2026-08-28T12:20:08.000Z"
        }
      ]
    },
    "cost": "0.750000",
    "settled": true,
    "createdAt": "2026-08-28T12:00:00Z",
    "completedAt": "2026-08-28T12:00:30Z"
  }
}

For each available asset, open or download assets[].url directly without an API Authorization header. The signed URL expires at expiresAt; poll the task again to refresh it while the asset remains within its retention period. A pending or unavailable asset has no URL. Treat signed URLs as private.

SDKs & integrations

Use SpicyAPI in your app, terminal, AI assistant or visual workflow. Choose a setup guide below.

SDKs

TypeScript · Python · Go · PHP · Java

CLI

Terminal commands and automation scripts.

MCP

Claude Code · Cursor · Codex

Agent Skill

Reusable instructions for your AI assistant.

n8n

Self-hosted community nodes and importable HTTP Request templates for n8n Cloud.

ComfyUI

Image, video and audio workflows on a node canvas.

More integrations

SillyTavern Telegram / Discord App proxy OpenAI SDK LangChain Vercel AI SDK

Browse ready-to-use workflows

You do not have to write any of the code above. Connect the MCP server once and any agent that speaks MCP — Claude Code, Codex, Cursor, Windsurf, VS Code, Gemini CLI, Cline, LibreChat — can run this model by name, or run it from your shell with a single command.

Say this to your agent

Use bytedance/omnihuman-2.0/talking-avatar to animate ./poster.png — slow push-in, five seconds — and save the clip next to it.

The agent looks the model up in the live catalogue, uploads the local file for you, then shows the exact quote and its maximum charge and waits for your confirmation. Nothing is billed until you say yes.

Connect it once

claude mcp add spicyapi -e SPICY_API_KEY=$SPICY_API_KEY \
  -- npx --yes --package=@spicyapi/mcp spicyapi-mcp
npx skills add spicyapi-ai/spicy-skill

The first line gives the agent the tools; the second gives it the Skill — the operating rules for them, so it quotes before it bills, uploads local files instead of pasting bytes, and reads the finished asset URL rather than guessing. Every other client takes the same server through its own config file: see the developer hub.

Or one command in your shell

npx @spicyapi/cli tasks create --model bytedance/omnihuman-2.0/talking-avatar \
  --input-json '{"audio_url":"https://example.com/reference.mp3","image_url":"https://example.com/reference.png","prompt":"A slow aerial shot drifting over a misty pine forest at sunrise"}' \
  --wait

The input is this endpoint’s own schema, same as the examples above. --wait polls until the task is finished and prints it; add --yes to skip the confirmation prompt in CI. Set SPICY_API_KEY first — one key works for the API, the CLI and the MCP server.

Local files

npx @spicyapi/cli files upload ./poster.png
# -> spicy://uploads/01k3m8x9q2z4v7n5p6r8s0t1w3

npx @spicyapi/cli tasks create --model bytedance/omnihuman-2.0/talking-avatar \
  --input-json '{"audio_url":"spicy://uploads/01k3m8x9q2z4v7n5p6r8s0t1w3"}' \
  --wait

Request bodies take URLs, not bytes. Upload the file first and pass the spicy:// URI it prints — the field name above is this endpoint’s own, straight from its schema. An agent on MCP does this step for you, which is why the sentence further up can name ./poster.png directly. Media already reachable over public HTTPS needs no upload.

How long results are kept

npx @spicyapi/cli tasks create --model bytedance/omnihuman-2.0/talking-avatar \
  --input-json '{…}' --retention 1h --wait

npx @spicyapi/cli tasks get $TASK_ID   # read back the retention that applied

--retention only ever shortens the window — 30m, 1h, 7d, or 0 to delete as soon as the task reaches a terminal state. The value that actually applies is the shortest of your request, your account settings and the platform maximum, so read it back from the task record rather than assuming the number you asked for took effect. Passing a longer window than your account allows does nothing.

Use it in ComfyUI

Install the SpicyAPI nodes from ComfyUI Manager or with the command below, then search for OmniHuman 2.0 · Talking Avatar on the canvas. The node has the same options as this page and shows its price before it runs.

cd ComfyUI/custom_nodes
git clone https://github.com/spicyapi-ai/comfyui-spicyapi

Workflow examples Setup guide Source on GitHub

The embedded panel reads the same schema as this page, so it cannot drift from the API. If your docs need the content to be searchable, read the JSON under /api/embed/models/ and render it yourself.

<iframe src="https://spicyapi.ai/embed/en/models/omnihuman-2-0/api?model=bytedance%2Fomnihuman-2.0%2Ftalking-avatar" width="100%" height="560" loading="lazy" title="Request example" style="border:1px solid rgba(0,0,0,.12)"></iframe>

Open the panel

Workbench

Send a request

API authentication

Input parameters

Track the task

Read the result

SDKs & integrations

Use it in ComfyUI