Model Hub
The Model Hub (also called Model Marketplace) is where you browse all available models, compare pricing, and get ready-to-use integration code.
Browsing Models
- Navigate to Model Hub in the sidebar (or Pricing)
- Browse the model card grid — each card shows the model name, provider, description, tags, and pricing
Searching and Filtering
- Search: Type a keyword in the search bar to filter by model name, description, or tags
- Provider Filter: Use the provider dropdown to show only models from a specific vendor (e.g., OpenAI, Anthropic, Google)
- Group Filter: Filter by access group if your account has group-based model access
- Sort: Sort models by name in ascending or descending order
Model Card Information
Each model card displays:
| Field | Description |
|---|---|
| Model Name | The identifier used in API requests (e.g., gpt-4o-mini) |
| Provider | The company that created the model |
| Description | A brief summary of the model's capabilities |
| Tags | Labels like vision, streaming, tools indicating supported features |
| Input Price | Cost per million input tokens |
| Output Price | Cost per million output tokens |
| Cache Price | Cost per million cached input tokens (if supported) |
Model Detail Page
Click any model card to open its detail page.
Overview
The detail page provides a comprehensive view of the model:
- Model name and provider at the top
- Tags showing supported capabilities
- Copy button to quickly copy the model name for use in your code
Description
A more detailed description of the model's capabilities, use cases, and limitations.
Pricing
The pricing section shows the exact per-million-token rates for:
- Input tokens — Tokens in your prompt
- Output tokens — Tokens in the model's response
- Cache tokens — Tokens served from cache (lower price, where applicable)
Integration Code
The detail page provides ready-to-use code examples in multiple languages:
- Python
- TypeScript
- Java
- Go
- cURL
You can also switch between API types:
- Chat Completions — For conversational models (most common)
- Text Completions — For legacy completion-style models
Copy the code, replace YOUR_API_KEY with your actual key, and you're ready to go.
Understanding Pricing
Model prices on TokenGO are calculated using a ratio system:
- Input Price is derived from the model's base ratio
- Output Price = Input Price x Completion Ratio (varies by model)
- Cache Price = Input Price x Cache Ratio (where supported)
All prices are displayed in USD per million tokens on the Model Hub, so you can directly compare costs across models.
Pricing Example
If a model shows:
- Input: $5.00/1M tokens
- Output: $15.00/1M tokens
Then a request using 1,000 input tokens and 500 output tokens would cost:
(1,000 / 1,000,000 x $5.00) + (500 / 1,000,000 x $15.00) = $0.005 + $0.0075 = $0.0125
Choosing the Right Model
| Use Case | Recommended Models |
|---|---|
| General chat & Q&A | GPT-4o mini, Claude 3.5 Haiku |
| Complex reasoning | GPT-4o, Claude 3.5 Sonnet, DeepSeek R1 |
| Code generation | Claude 3.5 Sonnet, GPT-4o |
| Long document processing | Gemini 1.5 Pro |
| Cost-sensitive applications | GPT-4o mini, Gemini 1.5 Flash |
| Vision / image understanding | GPT-4o, Claude 3.5 Sonnet |
Try any model in the Playground before integrating it into your application.