Search

PythonReactSEOPDF
OpenClaw

siliconflow-media

SiliconFlow 多模态服务,支持图片生成(FLUX/Qwen)、视频生成(Wan)、TTS语音合成、ASR语音识别。使用代金券支付。

axdlee0 source installs☆ 0 Repository stars

Installation

Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.

SkillHub CLI

skillhub install axdlee/siliconflow-media

Inspect the source and compatibility requirements, then follow its installation steps.

Give your agent the source link and ask it to inspect instructions and dependencies before explaining the installation steps.

View source ↗Guides

Skill instructions

SiliconFlow 媒体服务

SiliconFlow 提供丰富的 AI 模型服务,支持代金券支付(当前余额 3000+)。

环境变量

  • SILICONFLOW_API_KEY - SiliconFlow API Key

🎨 图片生成

uv run {baseDir}/scripts/image_gen.py --prompt "描述" --filename "output.png" [--model MODEL]

可用模型:

参数 模型 说明
flux (默认) FLUX.1-schnell 快速高质量,约 5-10 秒
flux-dev FLUX.1-dev 开发版
flux-pro FLUX.1-pro 专业版
qwen Qwen/Qwen-Image 通义图片生成
qwen-edit Qwen/Qwen-Image-Edit 图片编辑
qwen-edit-2509 Qwen/Qwen-Image-Edit-2509 最新编辑版

示例:

# FLUX 快速生成
uv run {baseDir}/scripts/image_gen.py --prompt "a cute robot assistant" --filename "robot.png"

# Qwen 生成
uv run {baseDir}/scripts/image_gen.py --prompt "山水画" --filename "landscape.png" --model qwen

🎬 视频生成

# 文生视频
uv run {baseDir}/scripts/video_gen.py --prompt "描述" --filename "output.mp4"

# 图生视频
uv run {baseDir}/scripts/video_gen.py --prompt "描述" --image "input.png" --filename "output.mp4"

模型:

  • 文生视频: Wan-AI/Wan2.2-T2V-A14B
  • 图生视频: Wan-AI/Wan2.2-I2V-A14B

⚠️ 视频生成时间较长(约 2-5 分钟)

🎤 语音合成 (TTS)

uv run {baseDir}/scripts/tts.py --text "要合成的文字" --filename "output.mp3" [--model MODEL]

可用模型:

参数 模型 说明
fish-speech (默认) fish-speech-1.5 Fish Audio 高质量
cosyvoice CosyVoice2-0.5B 阿里语音克隆
indextts IndexTTS-2 Index TTS
moss MOSS-TTSD-v0.5 MOSS 多语言

示例:

uv run {baseDir}/scripts/tts.py --text "你好世界" --filename "hello.mp3"

👂 语音识别 (ASR)

uv run {baseDir}/scripts/asr.py --audio "input.mp3" [--model MODEL]

可用模型:

参数 模型 说明
sensevoice (默认) SenseVoiceSmall 阿里语音识别
teleai TeleSpeechASR TeleAI 识别

示例:

uv run {b...

Source information

Source
github
Repository
openclaw/skills
Path
skills/axdlee/siliconflow-media
Version
—
Commit
1d802c1567a4eff68cf068441ffc7adcc09c876a
Classification source
rules
Status
active
Visibility
public

These signals come from indexed source analysis and may lag behind changes. Check the current source files.

Extracted content

SKILL.md
# SiliconFlow 媒体服务

SiliconFlow 提供丰富的 AI 模型服务,支持代金券支付(当前余额 3000+)。

## 环境变量

- `SILICONFLOW_API_KEY` - SiliconFlow API Key

## 🎨 图片生成

```bash
uv run {baseDir}/scripts/image_gen.py --prompt "描述" --filename "output.png" [--model MODEL]
```

**可用模型**:
| 参数 | 模型 | 说明 |
|------|------|------|
| `flux` (默认) | FLUX.1-schnell | 快速高质量,约 5-10 秒 |
| `flux-dev` | FLUX.1-dev | 开发版 |
| `flux-pro` | FLUX.1-pro | 专业版 |
| `qwen` | Qwen/Qwen-Image | 通义图片生成 |
| `qwen-edit` | Qwen/Qwen-Image-Edit | 图片编辑 |
| `qwen-edit-2509` | Qwen/Qwen-Image-Edit-2509 | 最新编辑版 |

**示例**:
```bash
# FLUX 快速生成
uv run {baseDir}/scripts/image_gen.py --prompt "a cute robot assistant" --filename "robot.png"

# Qwen 生成
uv run {baseDir}/scripts/image_gen.py --prompt "山水画" --filename "landscape.png" --model qwen
```

## 🎬 视频生成

```bash
# 文生视频
uv run {baseDir}/scripts/video_gen.py --prompt "描述" --filename "output.mp4"

# 图生视频
uv run {baseDir}/scripts/video_gen.py --prompt "描述" --image "input.png" --filename "output.mp4"
```

**模型**:
- 文生视频: `Wan-AI/Wan2.2-T2V-A14B`
- 图生视频: `Wan-AI/Wan2.2-I2V-A14B`

⚠️ 视频生成时间较长(约 2-5 分钟)

## 🎤 语音合成 (TTS)

```bash
uv run {baseDir}/scripts/tts.py --text "要合成的文字" --filename "output.mp3" [--model MODEL]
```

**可用模型**:
| 参数 | 模型 | 说明 |
|------|------|------|
| `fish-speech` (默认) | fish-speech-1.5 | Fish Audio 高质量 |
| `cosyvoice` | CosyVoice2-0.5B | 阿里语音克隆 |
| `indextts` | IndexTTS-2 | Index TTS |
| `moss` | MOSS-TTSD-v0.5 | MOSS 多语言 |

**示例**:
```bash
uv run {baseDir}/scripts/tts.py --text "你好世界" --filename "hello.mp3"
```

## 👂 语音识别 (ASR)

```bash
uv run {baseDir}/scripts/asr.py --audio "input.mp3" [--model MODEL]
```

**可用模型**:
| 参数 | 模型 | 说明 |
|------|------|------|
| `sensevoice` (默认) | SenseVoiceSmall | 阿里语音识别 |
| `teleai` | TeleSpeechASR | TeleAI 识别 |

**示例**:
```bash
uv run {b...

Related skills

OpenClaw

13-Day Sprint Method

galizki

Productivity system based on Maya calendar with 13 natural tones for project management and personal development

OpenClaw

1password

daggerkun007

Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in (single or multi-account), or reading/injecting/running secrets via op.