Add media models to the AI catalogue

[softmarshmallow]

Summary

• add provider-grounded video, sound-effect, and 3D model records
• surface supported media models and provider-native pricing on /ai/models
• add official maker marks across web and Desktop settings
• expose the media catalogue through the editor AI catalogue
• enforce that unsupported models never enter the video catalogue

Support boundary

Only models Grida can actually select and call enter the video catalogue. Gemini Omni and Seedance 2.5 are intentionally absent until their real execution and metering contracts are integrated.

Sound-effect and 3D records remain explicitly staged: they are available to the dedicated Desktop compatibility playground, but catalogue presence does not claim web generation support.

Review fixes

• removed Gemini Omni rather than representing fal's token-native meter as exact per-second pricing
• removed the compatibility-only README contradiction and made listed: false structurally invalid for video cards
• updated Grok Imagine Video 1.5 to stable Vercel/fal IDs, the grounded 1–15s duration range, 480p/720p/1080p prices, and its input-image surcharge
• fixed the public page's H1/H2/H3 document outline and marked adjacent maker logos decorative
• kept SFX and 3D entries visible across future staged → listed lifecycle changes
• kept ElevenLabs pricing and its official clear-space construction provider-native rather than inventing a normalized abstraction

Validation

• @grida/ai-models build
• @grida/ai-models tests: 64/64
• @grida/ai-models typecheck
• @grida/react-icons build and typecheck
• editor typecheck
• scoped oxfmt and oxlint
• git diff --check
• live /ai/models request: HTTP 200
• independent code, topology, accessibility, OSS, and security review

Reviewer notes

• Human decision before merge: Tencent's official media-resources page says its media-library assets are for personal or editorial use. Confirm that using the Tencent mark in the public catalogue and Desktop settings is acceptable, or replace/remove it.
• ElevenLabs' visually smaller “11” footprint preserves the official brand guide's 3× holding-device construction; it was intentionally not cropped or optically redrawn.
• The per-model SEO registry and reusable gate topology groundwork were deliberately moved to a separate stacked branch so this PR remains catalogue-only.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
grida	Ready	Preview	Aug 19, 2026 9:29am

5 Skipped Deployments

Project	Deployment	Actions	Updated (UTC)
backgrounds	Ignored	Preview	Aug 19, 2026 9:29am
blog	Ignored	Preview	Aug 19, 2026 9:29am
code	Ignored		Aug 19, 2026 9:29am
docs	Ignored	Preview	Aug 19, 2026 9:29am
viewer	Ignored	Preview	Aug 19, 2026 9:29am

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: d5085ff5-1ece-436d-bdfd-4e2013382db8

📥 Commits

Reviewing files that changed from the base of the PR and between 5909675 and 5abd34a.

📒 Files selected for processing (12)

• .agents/skills/ai-models/SKILL.md
• editor/app/(www)/(ai)/ai/models/page.tsx
• editor/app/desktop/settings/page.tsx
• editor/lib/ai/ai.ts
• packages/grida-ai-models/README.md
• packages/grida-ai-models/__tests__/models.test.ts
• packages/grida-ai-models/src/models.ts
• packages/grida-react-icons/src/logos/bytedance.tsx
• packages/grida-react-icons/src/logos/elevenlabs.tsx
• packages/grida-react-icons/src/logos/index.ts
• packages/grida-react-icons/src/logos/microsoft.tsx
• packages/grida-react-icons/src/logos/tencent.tsx

Included review availability: Your plan provides up to 8 included reviews per hour; 7 remain after this review.

---

Walkthrough

The PR expands AI model catalogue contracts and documentation, adds video pricing with input-image surcharges, exports video and 3D catalogues, adds provider logos, and extends the models page with video, sound-effect, and 3D sections.

Changes

AI model catalogue and pricing

Layer / File(s)	Summary
Provider logo components and mappings packages/grida-react-icons/src/logos/*, editor/app/desktop/settings/page.tsx	Adds ByteDance, ElevenLabs, Microsoft, and Tencent logos. Registers the logos for model vendors and the ElevenLabs BYOK provider.
Video catalogue pricing and enablement packages/grida-ai-models/src/models.ts, .agents/skills/ai-models/SKILL.md, packages/grida-ai-models/README.md, packages/grida-ai-models/__tests__/models.test.ts	Adds optional input-image pricing, requires listed runtime-enabled video cards, updates Grok pricing and duration data, and extends catalogue tests and guidance.
Catalogue API exports editor/lib/ai/ai.ts	Exports the three_d and video catalogues through the ai namespace.
Models page sections and rendering editor/app/(www)/(ai)/ai/models/page.tsx	Adds shared vendor-logo rendering and sections for video, sound-effect, and 3D models. Displays formats, outputs, status, provider pricing, and input-image surcharges.

Estimated code review effort: 4 (Complex) | ~45 minutes

Merge Risk: ⚪ Minimal · up to 5abd3

The catalogue and presentation updates have no supplied concrete correctness, security, availability, or deployment risk at the current head, so no actionable merge-blocking risk remains after normal checks.

Sequence Diagram(s)

sequenceDiagram
  participant ModelsPage
  participant ai.video
  participant VideoCatalogue
  participant ProviderBindings
  ModelsPage->>ai.video: listed_models()
  ai.video->>VideoCatalogue: read enabled video cards
  VideoCatalogue->>ProviderBindings: resolve provider pricing
  ProviderBindings-->>ModelsPage: return rates and input-image surcharges

Loading

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly summarizes the main change: adding media models to the AI catalogue.
Description check	✅ Passed	The description directly explains the catalogue, pricing, provider logos, support boundaries, and validation for the changeset.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch chore/ai-model-catalogue-media

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5abd34a199

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Stop publishing ungrounded OpenRouter video rates

When visitors view Veo or Seedance on /ai/models, this loop renders their OpenRouter bindings as per-second prices, even though .agents/skills/ai-models/SKILL.md explicitly states that OpenRouter exposes only $0/MTok for video and has no usable per-second meter, and instructs maintainers to leave such bindings out. Consequently the page presents $0.400/s and $0.067/s as “grounded native pricing” without a supported meter; omit these bindings from the catalogue/display until a verifiable per-second rate exists.

Useful? React with 👍 / 👎.