# CogVLM2 Video

> Video understanding and captioning — ask free-form questions about a video and get detailed answers about actions, scenes, and content.

- **Provider**: replicate
- **Model ID**: chenxwh/cogvlm2-video
- **Category**: image_video_caption
- **Credits**: 28 per request
- **Pricing Type**: fixed

## API Endpoint

Base URL: https://api.core.today/v1

### Create Prediction
POST /predictions

### Get Status
GET /predictions/{job_id}

### Cancel
DELETE /predictions/{job_id}

## Authentication

Header: `X-API-Key: YOUR_API_KEY`

## Input Parameters

- `top_p` (number, optional): When decoding text, samples from the top p percentage of most likely tokens; lower to ignore less likely tokens (Default: `0.1`; Range: min: 0, max: 1)
- `temperature` (number, optional): Adjusts randomness of outputs, greater than 1 is random and 0 is deterministic (Default: `0.1`; Range: min: 0)
- `input_video` (string, **required**): Input video
- `prompt` (string, optional): Input prompt (Default: `Describe this video.`)
- `max_new_tokens` (integer, optional): Maximum number of tokens to generate. A word is generally 2-3 tokens (Default: `2048`; Range: min: 0)

## Example Request

```json
{
  "model": "chenxwh/cogvlm2-video",
  "input": {
    "top_p": 0.1,
    "temperature": 0.1,
    "input_video": "https://replicate.delivery/pbxt/LgGGFqrlw37TQMMWbjCTvYwp1vhH916HGqjKIuLxyB5SqiuT/%E5%A4%A7%E8%B1%A1.mp4",
    "prompt": "\u8bf7\u4ed4\u7ec6\u63cf\u8ff0\u8fd9\u4e2a\u89c6\u9891",
    "max_new_tokens": 2048
  }
}
```

## Response Format

```json
{
  "job_id": "abc123",
  "status": "pending",
  "provider": "replicate",
  "model": "black-forest-labs/flux-schnell",
  "created_at": "2026-01-01T00:00:00Z",
  "result": null,
  "error": null
}
```

Status values: `pending`, `processing`, `completed`, `failed`, `cancelled`

## Usage Flow

1. POST /predictions with model and input → receive job_id
2. Poll GET /predictions/{job_id} until status is `completed` or `failed`
3. Result contains output URL(s) or data

## Output Type

json

## Tags

video-captioning, video-understanding, vqa, cogvlm

## Documentation

https://replicate.com/chenxwh/cogvlm2-video