[Runyour Guide #4] Selecting and calling your preferred model in the API Router

Hi, this is the Runyour team.

Have you ever struggled with choosing an AI model?

"Which model is best for this task?"

"It performs well, but how much will it cost?"

"How large should the context window be for long documents?"

The more options you have, the harder it can be to choose. To help with this process, the Runyour API router has improved the way you explore and select models.

In this guide, we’ll show you how to find the perfect model for your project using our sorting, filtering, and purpose-based recommendation features.

🎬 Follow along with the video

Watch the video to see the entire process, from browsing the model list and sorting by criteria to getting recommendations and generating your API key.

What you'll learn in the video:

STEP 1. Explore supported models

You can compare the input/output costs, context window size, and maximum output for each model on a single screen. The exact model ID required for API calls is also displayed, so you can simply copy and paste it.

STEP 2. Sort and filter by your criteria

There's no need to scroll through a long list; you can reorder the models based on the metrics that matter most to you.

Sort by context size : When you need to handle long documents or large codebases.
Price : Useful when narrowing down models to fit your budget or selecting a model for high-traffic features.
By Provider · By Type : Useful when you want to view models from a specific provider or filter by text and image models.

You can switch between list and card views to browse in the way that works best for you.

STEP 3. Get model recommendations for your needs

If you're not sure which model is right for you, instead of comparing specs, start with your goal and go from there.

Open [Recommend a model for me], select what you want to build, and it will suggest the best-suited models.

Writing & Drafting — Marketing copy, emails, and report drafts
Summarizing & Organizing — Extracting tasks from long meeting minutes, PDFs, and tables
Image Generation — Ads, social media, and product images
Classification & Extraction — Categorizing inquiries and extracting specific information
Adding AI to your service — Integrating chat and generation features into your own service
Agents & Automated Workflows — Automated flows that call tools in sequence

Recommended models include labels like 'Natural long-form writing', 'Balanced quality and cost', or 'Fast draft iteration', explaining why it was chosen, making it easy to decide even if you aren't familiar with the technical specs.

Settings menus and item names may vary depending on the Kilo Code version, so we recommend watching the video guide if you are connecting for the first time.

STEP 4. Create and call an API Key

Once you have found a model you like, create an API Key in the [API Router] menu.

‍Call all 159 models with a single API Key You can use the same key and settings even if you switch models. Since it supports an OpenAI-compatible interface, you can switch to a different model simply by changing the model value while keeping your existing call structure.

Another advantage of the API Router is that it removes the burden of comparing multiple models to find the one that fits best.