Models Catalog
Endpoint-first navigation for production work. Each modality reference lists curated picks organized by use case (premium realism / fast & cheap / 4K / specialized). Before reaching for free-text search, consult the modality reference that matches the task.
Runtime: All endpoint calls use the vg generate CLI (npm install -g vibedgames, or pnpm dogfood in this repo). The API key lives on the vibedgames server, so there is no per-machine setup. See the generate skill for the command reference.
Endpoint-first rule
- Pick the endpoint ID from the right modality reference.
- Verify it:
vg generate models --endpoint_id <endpoint_id> --json.
- Inspect it:
vg generate schema <endpoint_id> --json.
- Check cost when relevant:
vg generate pricing <endpoint_id> --json.
- Use text search only if the routed endpoint is missing, deprecated, rejected, or the role is not covered here:
vg generate models "<task description>" --json
vg generate docs "<topic>" --json
Do not invent endpoint IDs.
Modality references
Load the reference matching the user's task:
- text-to-image.md, image generation from prompt (text-heavy, premium still, fast draft)
- image-to-image.md, image editing, inpainting, background removal, upscaling
- text-to-video.md, video generation from prompt (highest quality, fast/economical, multi-shot storytelling)
- image-to-video.md, video from a reference frame (including audio-driven and lip-sync variants)
- video-to-video.md, video edit, restyle, upscale, background removal
- text-to-3d.md, 3D model generation from text
- image-to-3d.md, 3D model generation from reference images
- text-to-audio.md. TTS, music, SFX generation
- audio-to-text.md, speech-to-text (Whisper, ElevenLabs Scribe with diarization)
- image-to-text.md. OCR, captioning, VQA, detection, segmentation
- campaign-ugc.md, use-case routing for marketing campaign assets and creator/UGC ads
Utility endpoints
For workflow utility endpoints (resize, composite, mask, audio merge, subtitle, etc.), search the catalog with vg generate models --category <modality> and inspect the candidates with vg generate schema <endpoint_id>.
Utility endpoints are explicit because they are deterministic tools, not creative model choices. Always inspect schema before use.
1---2name: model-catalog3description: Pick a `vg generate` endpoint for a task, and the default endpoint IDs the media skills use when none is named.4---56# Models Catalog78Endpoint-first navigation for production work. Each modality reference lists curated picks organized by use case (premium realism / fast & cheap / 4K / specialized). Before reaching for free-text search, consult the modality reference that matches the task.910> **Runtime:** All endpoint calls use the `vg generate` CLI (`npm install -g vibedgames`, or `pnpm dogfood` in this repo). The API key lives on the vibedgames server, so there is no per-machine setup. See the `generate` skill for the command reference.1112## Endpoint-first rule13141. Pick the endpoint ID from the right modality reference.152. Verify it: `vg generate models --endpoint_id <endpoint_id> --json`.163. Inspect it: `vg generate schema <endpoint_id> --json`.174. Check cost when relevant: `vg generate pricing <endpoint_id> --json`.185. Use text search only if the routed endpoint is missing, deprecated, rejected, or the role is not covered here:1920```bash21vg generate models "<task description>" --json22vg generate docs "<topic>" --json23```2425Do not invent endpoint IDs.2627## Modality references2829Load the reference matching the user's task:3031- [text-to-image.md](references/text-to-image.md), image generation from prompt (text-heavy, premium still, fast draft)32- [image-to-image.md](references/image-to-image.md), image editing, inpainting, background removal, upscaling33- [text-to-video.md](references/text-to-video.md), video generation from prompt (highest quality, fast/economical, multi-shot storytelling)34- [image-to-video.md](references/image-to-video.md), video from a reference frame (including audio-driven and lip-sync variants)35- [video-to-video.md](references/video-to-video.md), video edit, restyle, upscale, background removal36- [text-to-3d.md](references/text-to-3d.md), 3D model generation from text37- [image-to-3d.md](references/image-to-3d.md), 3D model generation from reference images38- [text-to-audio.md](references/text-to-audio.md). TTS, music, SFX generation39- [audio-to-text.md](references/audio-to-text.md), speech-to-text (Whisper, ElevenLabs Scribe with diarization)40- [image-to-text.md](references/image-to-text.md). OCR, captioning, VQA, detection, segmentation41- [campaign-ugc.md](references/campaign-ugc.md), use-case routing for marketing campaign assets and creator/UGC ads4243## Utility endpoints4445For workflow utility endpoints (resize, composite, mask, audio merge, subtitle, etc.), search the catalog with `vg generate models --category <modality>` and inspect the candidates with `vg generate schema <endpoint_id>`.4647Utility endpoints are explicit because they are deterministic tools, not creative model choices. Always inspect schema before use.
kyh