The Skytells Console: Run 50+ AI Models Without Writing a Line of Code

The Skytells Console: Run 50+ AI Models Without Writing a Line of Code

There's a particular kind of friction that kills adoption of any developer tool: the gap between "I want to try this" and actually seeing a result. With most AI platforms, that gap involves reading documentation, installing an SDK, configuring authentication, writing a script, debugging a script, and then — maybe — seeing output.

The Skytells Console eliminates that gap entirely.

It's an interactive, browser-based playground where you can select any model from the Skytells catalog, fill in parameters through a visual form, click Run, and see results immediately. Generate an image. Create a video from a text prompt. Lip-sync footage in 40+ languages. Compose a song. Transcribe audio. All from the same interface, all without writing code.

But here's the part that makes it genuinely useful rather than just convenient: it generates working code for you. Every interaction you run in the Console produces copy-paste-ready snippets in Node.js, Python, and cURL. When you're ready to move from experimentation to integration, the code is already written.

What the Console Looks Like

The interface is a two-panel layout designed around a simple workflow: configure on the left, see results on the right.

The left panel has your input controls. When you select a model, the Console reads its input schema and auto-generates the appropriate form fields — text areas for prompts, sliders for numeric parameters like steps or guidance scale, toggles for boolean options, dropdowns for enum choices, and drag-and-drop fields for image uploads. You don't read documentation to figure out what parameters a model accepts. The form shows you.

Below the form, tabs switch between the visual editor, raw JSON input, and auto-generated code in Node.js, Python, and HTTP/cURL. This means you can prototype a request visually, then switch to the code tab and grab the exact API call that reproduces it.

The right panel shows output. Images render inline with lightbox preview. Videos play with full controls. Audio gets a custom player with seek, progress, and volume. And there's a JSON tab showing the raw response, plus a Logs tab with timestamped request/response details for debugging.

The Model Catalog

The Console connects to the full Skytells model catalog — currently over 50 models spanning every major AI modality:

Image generation. TrueFusion Large, TrueFusion Pro, and TrueFusion Pano cover everything from standard image generation to panoramic wide-angle shots. Type a prompt, adjust parameters with sliders, click Run, see a hyper-realistic image in seconds. FID scores of 1.8 to 2.3 — that's photo-grade quality.

Video generation. TrueFusion Video and TrueFusion Video Pro for text-to-video generation — describe a scene and get a fully rendered video clip. The platform also supports third-party video models like Sora 2. For lip-sync and video dubbing, LipFusion re-renders speaker footage to match new audio with 98.2% sync accuracy across 40+ languages. The Console handles async processing automatically — you don't need to poll for results or manage callbacks.

Audio and music. BeatFusion and BeatFusion Pro for text-to-music generation. Write lyrics with section tags, describe a style, and get back a full song — vocals, instrumentation, mixing — in broadcast quality. The audio player in the output panel lets you listen immediately.

Language models. DeepBrain for advanced reasoning and text generation. The Console streams responses in real-time, so you see tokens appear as they're generated.

And more. Speech-to-text, text-to-speech, image editing, video-to-video transformation — every model in the catalog is playable through the same interface. Filter by capability using the chips at the top: Text to Image, Text to Video, Audio, Fast, and a dozen other categories.

No-Code Doesn't Mean No Power

The term "no-code" sometimes implies "limited." Not here.

Every parameter that's available through the API is available in the Console. Guidance scale, inference steps, resolution, aspect ratio, temperature, max tokens, seed values — the form auto-generates controls for all of them. Advanced users get the same level of control they'd have writing raw API calls. Beginners get sensible defaults and visual feedback.

The model header for each model displays exactly what you need to know before running a prediction: pricing per unit, hardware (H100 GPUs), current status, cold boot time, and whether the model is official, partner-provided, or open-source. No surprises on the bill.

From Console to Code in One Tab Switch

This is the feature that turns the Console from a demo tool into a development accelerator.

Say you've been generating images in the Console, tweaking prompts and adjusting guidance scale until the output is exactly what you want. Click the "Node.js" tab in the left panel, and you see something like this:

import { createClient } from 'skytells';

const skytells = createClient("YOUR_API_KEY");

const result = await skytells.predict({
  model: "truefusion-pro",
  input: {
    prompt: "Product photography, minimalist white background, professional lighting, luxury watch on marble surface",
    steps: 18,
    guidanceScale: 7.5
  },
  await: true
});

That's a working code snippet with your exact parameters already filled in. Copy it into your project, replace the API key placeholder with your actual key from the Dashboard, and you have production code.

Same for Python:

import requests

response = requests.post(
    'https://api.skytells.ai/v1/predict',
    headers={
        'Content-Type': 'application/json',
        'x-api-key': 'YOUR_API_KEY'
    },
    json={
        'model': 'truefusion-pro',
        'input': {
            'prompt': 'Product photography, minimalist white background...',
            'steps': 18,
            'guidanceScale': 7.5
        },
        'await': True
    }
)

The code generation is one of those features that seems minor until you use it. It eliminates the translation step between "I know what I want" and "I know how to ask for it programmatically."

Prediction History

Every prediction you run in the Console is saved and accessible at /console/predictions. Each entry shows:

• The model used
• Input parameters
• Output (with visual preview for images, video, and audio)
• Processing duration
• Hardware allocated
• Cost

This is useful for comparing results across different parameter configurations, tracking your usage, and going back to find that one generation from last Tuesday that you forgot to download. It's also the fastest way to see your spending patterns and optimize costs.

Who This Is For

The Console serves three distinct groups, and it works slightly differently for each:

Non-technical teams. Marketing, content, and design teams who need AI-generated assets but don't have (or want) a development workflow. They open the Console, select a model, describe what they want, and get results. No code, no CLI, no API keys to manage manually. The Console handles authentication through your Skytells account.

Developers exploring the API. Before writing integration code, you want to understand what a model can do, what parameters it accepts, and what the output looks like. The Console is faster than reading docs because you see actual results for actual inputs. When you're satisfied, the code tab gives you a working starting point.

Product teams evaluating models. If you're deciding which model to use for a feature — TrueFusion Video vs. Video Pro vs. Sora 2, TrueFusion Large vs. Pro for images, BeatFusion Standard vs. Pro — the Console lets you run the same input through different models side by side and compare output quality, generation speed, and cost. No integration work required for evaluation.

Real-Time Cost Visibility

One of the consistent complaints about AI APIs is bill shock — you experiment for an afternoon and discover you've spent $200. The Console addresses this directly.

Every model shows its pricing in the model header: cost per image, per second of video, per token, or whatever the relevant unit is. Before you click Run, you know what it costs. After you run it, the prediction detail page shows the exact charge.

You can also set budget alerts and hard spending caps through the Dashboard. The Console respects these limits — if you're approaching your cap, you'll know before you hit it.

Getting Started

The Console is live at skytells.ai/console. Sign in with your Skytells account (or create one — it takes 30 seconds), and you're in. Your API key is automatically populated in the console, so there's zero configuration.

Pick a model. Fill in the form. Click Run.

If you're coming from the API side and want to understand what models are available, the Console is the fastest way to browse the catalog. If you're coming from the business side and want to generate content without a developer in the loop, the Console is the product. Either way, you get results in under a minute.

The full API documentation is at docs.skytells.ai when you're ready to integrate. The Skytells Learn platform has guided tutorials for every model and modality. And the Console will be here whenever you want to try something new without writing code first.