Model Card
The single-model view in the Skytells Console — pricing, SDK snippet, capability tags, playground access, and API schema in one place.
The Model Card is the detail page for a specific model. Open it by clicking any model row in the Model Catalog. It consolidates everything you need to evaluate and integrate a model: what it does, how much it costs, and a direct path into the Playground for live testing.
Accessing a model card
In the Skytells Console, go to AI → Models, then click any row in the catalog. The model card opens at /ai/models/{vendor}/{model}.
You can also reach a model card directly from a Playground session or from a prediction detail.
Model card layout
Header
The header identifies the model at a glance:
| Element | Description |
|---|---|
| Provider logo | Visual identifier for the model's provider (Skytells, Black Forest Labs, OpenAI, etc.). |
| Model name | Full name, e.g., Black Forest Labs / Flux 2 Flex or TrueFusion. |
| Namespace | The identifier you pass to the API — e.g., skytells/truefusion, black-forest-labs/flux-2-flex. Copyable with one click. |
| Status badges | Official (published by provider or Skytells), Operational (currently available), output type (Image, Video, Audio, Text). |
| Try in Playground | Opens the Inference Playground pre-loaded with this model. |
| Skytells Docs | Link to deeper API documentation for this model (opens in new tab). |
Status badges explained
| Badge | Meaning |
|---|---|
Official | This model is officially published and verified by its provider or by Skytells. |
Operational | The model is live and accepting requests. No active incidents. |
Image / Video / Audio / Text | The primary output type of this model. |
Multiple Pricing | The model has more than one pricing tier or pricing varies by resolution, quality, or other parameter. |
H100 | This model runs on NVIDIA H100 hardware — high-performance, enterprise-grade inference. |
Public | This model is accessible to all Skytells users without additional approval. |
Description
A short description of the model's purpose and design philosophy, sourced from the model provider. May include:
- Primary use case (e.g., "Max-quality image generation and editing with support for ten reference images").
- Architecture notes.
- Capability tags (e.g.,
text-to-image,image-editing,reference-image).
Skytells SDK panel
Every model card includes a ready-to-use snippet pre-filled with the correct model namespace. Copy it directly into your project — see TypeScript SDK — Predictions for the full client.run() interface.
Pricing table
The pricing table shows all pricing tiers for this model:
| Column | Description |
|---|---|
| Rule | The pricing rule name (e.g., Base pricing, HD tier, Reference images). |
| Condition | Any condition that must be met for this tier to apply. — means it applies unconditionally. |
| Rate | The cost per unit — e.g., $0.03 / image, $0.196 / second, $0.50 / 1M tokens. |
Models with Multiple Pricing badges have more than one row in this table, typically reflecting quality, resolution, or output length tiers.
Playground tab
The Playground tab within the model card is the full interactive testing surface for this model. See Inference Playground for a complete reference of the playground UI.
The playground tab has two panels:
Input panel (left):
- Form view — structured fields for each model parameter (Prompt, Input Images, Aspect Ratio, Resolution, etc.).
- JSON view — raw JSON input for advanced users or when pasting schemas.
- Node.js, Python, HTTP — generated code snippets for the current parameter values. Copy and paste directly into your application.
Output panel (right):
- Preview — rendered output (image displayed inline, video playable, audio playable).
- JSON — the Prediction object from the Predictions API.
- Logs — build and runtime output from the prediction execution.
API tab
The API tab shows the full request schema for this model — every accepted parameter, its type, whether it's required, and its description.
Examples tab
The Examples tab shows curated prompt examples and expected outputs for this model — useful for understanding the model's behavior and getting your first successful result quickly.
README tab
The README tab contains the model provider's original documentation — architecture notes, best practices, limitations, and advanced usage patterns.
Common workflows
Evaluate a model before integrating
- Open the model card from the catalog.
- Review the description, capability tags, and pricing.
- Select Try in Playground.
- Enter a representative input.
- Select Run and inspect the output.
- Switch to the Node.js or Python tab to copy the integration snippet.
Understand the input contract
- Open the model card.
- Go to the API tab.
- Review the full input schema — required vs. optional fields, accepted types, enumerations, and constraints.
Get example prompts
- Open the model card.
- Go to the Examples tab.
- Browse curated prompt-output pairs for this model.
Related
- Model Catalog — browse and filter all models.
- Inference Playground — the full interactive testing surface.
- Predictions — every run from the Playground appears here.
- API: Get Model — fetch model metadata programmatically.
- TypeScript SDK — Predictions —
client.run()andclient.predict()reference.
How is this guide?
Model Catalog
Browse and search the full catalog of Skytells inference models — filter by type, provider, and capability. Open any model to view pricing, SDK snippets, and input schema.
Predictions
A live, unified log of every inference request across your API keys, SDKs, CLI, and Playgrounds — with status, latency, cost, source, and full output inspection.