Tool catalog
19 tools. All available to all users. Version: 2026-04-19
When to use
Use this to scope your agent surface
Use the catalog to understand which tools are available, their sync/async behavior, and which providers each tool supports. Always call tools/list at runtime for the canonical schemas.
| Tool | Category | Async | Providers | Description |
|---|---|---|---|---|
| analyze_media | analysis | mixed | anthropic, grok, openai, qwen, soundside.ai, vertex | Technical ffprobe analysis, AI vision QA, transcription, segment detection, and EDL export. |
| apply_effect | editing | sync | soundside.ai | Apply Ken Burns, speed ramp, film grain, or vignette effects. |
| compose_media | editing | sync | soundside.ai | Add text, overlay media, or build split-screen composites. |
| compose_video | composition | sync | soundside.ai | Server-side video composition pipeline: enrich plan, generate assets in parallel, assemble with transitions, audio ducking, and overlays. |
| create_artifact | generation | sync | docx, gamma, mermaid, plotly, pptx, weasyprint | Create charts, presentations, documents, and diagrams. |
| create_audio | generation | mixed | minimax, runway, vertex | Text-to-speech, voice cloning, voice design, and sound effects. |
| create_image | generation | mixed | alibaba, grok, luma, minimax, runway, vertex | Generate images from text prompts with character reference support. |
| create_music | generation | async | minimax | Generate music from lyrics and style prompts. |
| create_text | generation | sync | grok, minimax, vertex | LLM chat completions with optional structured JSON output. |
| create_video | generation | async | alibaba, grok, luma, minimax, runway, vertex | Generate video from text or image. Supports I2V, video extension, and character references. |
| edit_audio | editing | sync | soundside.ai | Mix, replace, or pad audio on existing media. |
| edit_video | editing | sync | soundside.ai | Core video transforms: trim, concat, crossfade, speed, loop, color grade, burn subtitles, and custom FFmpeg. |
| extract_media | editing | sync | soundside.ai | Extract frames, frame sets, or audio tracks from media. |
| lib_list | library | sync | soundside.ai | Browse projects, collections, resources, lineage, and brand kits (free). |
| lib_manage | library | sync | soundside.ai | CRUD for projects, collections, resources, and brand kits. |
| lib_share | library | sync | soundside.ai | Share projects with other users by email with permission levels. |
| list_adapters | adapters | sync | soundside.ai | List LoRA adapters mirrored into the Soundside library (free). |
| manage_adapter | adapters | sync | soundside.ai | Inspect, deploy, undeploy, delete, or select a checkpoint for a LoRA adapter. |
| train_adapter | adapters | async | dashscope, modal | Train a LoRA adapter from library media on DashScope (Wan) or Modal (Hunyuan/LTX) backends. |
Pricing
Live pricing
Soundside aims to break even on provider pass-through costs with a minimal margin (~10%). Editing and library tools are $0.01/call. `analyze_media` is mode-priced: $0.01 for `technical` and `export_edl`, $0.02 for `transcribe` and `detect_segments`, and $0.03 for `vision_qa`.
Machine-readable pricing is always available at: GET /api/x402/status
Prices are DB-driven and may change without redeployment. Always check the endpoint rather than hardcoding.
Developer docs
Detailed tool reference
For full parameter documentation, examples, and tips, see the Tool Reference on GitHub.
Operational notes
Legacy references
- • creative_freedom is a self-hosted GPU provider available via API key only; it is intentionally omitted from the public x402 catalog.
- • batch_dispatch and batch_wait are internal runtime concerns, not user-facing MCP tools.