Introduction to EidoStack

EidoStack is a Multi-Model AI Playground for Developers that brings AI models from different providers into one interface.

Chat with AI models, solve technical problems, experiment with prompts, compare responses, and explore token usage, context, and request costs.

Connect your own OpenAI, Anthropic, or Google AI API keys and start using supported models directly in your browser—no additional software required.

This guide introduces EidoStack, explains what you can do in the Playground, and shows you how to get started.

What is EidoStack?

EidoStack is a browser-based platform for interacting with AI models, testing them, and exploring their capabilities.

Different AI models have different strengths, characteristics, and limitations. They may interpret instructions, solve technical problems, generate code, and handle conversation context in different ways.

EidoStack brings models from multiple providers into one workspace so you can explore their capabilities firsthand and use the models that work best for your tasks.

Unlike a basic AI chat, the Playground gives you additional tools for working with models:

• Select AI models from different providers.
• Compare responses from multiple models.
• Configure system prompts.
• Manage conversation context.
• Analyze input and output tokens.
• Track estimated API request costs.
• Visualize context-window usage.
• Save and organize chats and experiment history.

Use EidoStack as a regular AI chat for everyday tasks or as a workspace for exploring model behavior in greater depth.

One Playground. Multiple AI models. More control over your experiments.

Who is EidoStack for?

EidoStack is designed for developers and anyone who regularly works with AI models and wants to understand their capabilities better.

Software developers

Use different models to generate and review code, find bugs, refactor implementations, write SQL, explore technologies, and solve engineering problems.

AI engineers and LLM developers

Test model behavior, experiment with system prompts and context, compare results, and analyze request costs while building AI-powered features.

Prompt engineers

Explore how prompt wording, system instructions, and supplied context affect responses from different models.

AI startups and technical teams

Explore new models, validate ideas, find suitable approaches for your products, and estimate AI usage costs.

AI consultants

Test models on client tasks, compare different approaches, and demonstrate how their capabilities differ.

EidoStack is also useful for anyone who wants to work with AI models from multiple providers in one interface instead of constantly switching between separate platforms.

What can you do with EidoStack?

Chat with AI models

EidoStack provides a familiar chat interface for interacting with your selected AI model.

Ask questions, send instructions, continue conversations, and use models for everyday or professional tasks.

The Playground supports models from OpenAI, Anthropic, and Google AI through your own API keys.

Instead of using a separate interface for each provider, you can work with supported models in one workspace.

Explore different AI model capabilities

Every AI model has its own characteristics, capabilities, and limitations.

With EidoStack, you can try different models on the same or different tasks, explore their response styles, assess how well they follow instructions, and examine their approaches to solving problems.

For example, you can ask several models to review the same TypeScript component and see which issues they identify and which fixes they suggest.

This is particularly useful when new models are released and you want to experience their capabilities yourself instead of relying solely on provider descriptions or public benchmarks.

Available features and input types depend on the selected model and what the Playground supports.

Compare AI model responses

Comparison mode lets you send the same request to selected models and view their responses side by side.

You can compare:

• Response content and quality.
• How accurately models follow instructions.
• Different approaches to the same task.
• Level of detail and formatting.
• Token consumption.
• Estimated cost of each request.

For example, one model may produce a more detailed technical solution while another provides a shorter, more economical answer.

EidoStack makes these differences visible in one interface.

Comparison is optional. You can work with a single model and turn on comparison only when you need it.

Experiment with prompts

Prompt wording and system instructions directly affect AI model behavior.

EidoStack lets you adjust prompts, configure system instructions, and observe how different models respond to the same requirements.

Use a system prompt to define a model's role, communication style, response format, and additional rules.

For example, you might instruct a model to act as an experienced TypeScript developer, return strict JSON, or provide concise technical explanations.

By adjusting instructions and trying requests again, you can discover which formulations work best for your task.

Manage conversation context

Context determines what information an AI model receives with the current request.

EidoStack lets you choose whether to send only the current prompt, include a system instruction without history, or add conversation messages to the model request.

When you include conversation history, you can send the full conversation or a selected portion of it.

You can also inspect context-window usage to understand how many tokens are consumed by system instructions, conversation history, and the current request.

This helps you explore how context affects model responses while keeping token usage under control.

Analyze tokens and request costs

Different AI models may use different numbers of tokens to perform the same task.

EidoStack displays usage statistics alongside model responses.

You can track:

• Input tokens sent to a model.
• Output tokens used to generate a response.
• Estimated API request cost.
• Context-window consumption.
• Usage statistics by model, chat, and provider.
• Changes in token consumption and costs over a selected period.

These insights help you understand the cost of working with AI models and compare the economics of different experiments.

Displayed costs are estimates. Actual charges depend on the relevant AI provider's pricing.

Save and organize experiments

Working with AI rarely ends with a single request.

You may revise prompts, switch models, continue conversations, and revisit previous results.

EidoStack saves chat history, model responses, and their associated usage statistics.

Create separate chats for different tasks and organize them into folders.

For example, you can structure your workspace by project, model, technical task, or prompt experiment.

Your Playground becomes a persistent workspace where you can return to your work instead of starting from scratch each time.

How to use EidoStack

You can get started with the Playground in a few steps.

Step 1. Choose an AI model

Open EidoStack and select a model from one of your connected providers.

Choose a familiar model or try a new one whose capabilities you want to explore.

Step 2. Send your request

Enter a prompt, ask a question, or describe the task you want to solve.

For example, ask a model to review code, explain a technical concept, write an SQL query, or suggest a solution to a problem.

Step 3. Explore the result

Read the response and assess how well it fits your task.

Consider its content, instruction following, token usage, and estimated request cost.

Step 4. Experiment with settings

Try changing the prompt, configuring a system instruction, or choosing a different context strategy.

See how those changes affect the response.

Step 5. Try another model

Switch to another supported model or enable comparison mode to send the same request to selected models.

Explore how their responses differ.

Step 6. Analyze usage

Review token counts, request costs, and context-window usage.

Your saved history lets you return to previous experiments and continue working.

You can repeat this process for any task where you want to use or explore AI models.

API key security and privacy

EidoStack follows the Bring Your Own Key (BYOK) approach: you connect your own AI provider API keys.

During standard Playground interactions, your API keys are stored in encrypted form directly in your browser.

The vault uses the Web Crypto API, and access is protected by a master password.

Your master password is not sent to the EidoStack server. Decrypted API keys become available in browser memory after you unlock the vault.

In standard Playground mode, requests are sent directly to the selected AI providers using your own API keys.

This means you retain control over your provider accounts, usage limits, and API costs.

The selected provider receives the content necessary to process each request. Consider that provider's terms and privacy practices before sending sensitive information.

EidoStack stores workspace data, including chats, messages, folders, settings, and usage statistics, so you can return to previous work.

To learn more, see:

• API key vault.
• Manage your data and delete your account.

How is EidoStack different from a regular AI chat?

A regular AI chat primarily provides an interface for talking to a model.

EidoStack also offers a full chat experience, but adds tools for exploring and testing AI models.

In one Playground, you can work with models from multiple providers, change system instructions, manage context, compare results, and analyze token consumption and request costs.

You do not need to constantly switch between provider platforms or manually combine usage statistics from different models.

EidoStack brings chatting, experimentation, and analysis together in one workspace.

What EidoStack does not do

EidoStack does not train or fine-tune AI models, change their internal parameters or weights, or create new models from your conversations.

The Playground is designed for hands-on interaction with existing AI models and exploration of their capabilities.

It is not an automated benchmarking system or a production application monitoring platform.

EidoStack helps you manually test models, explore their behavior, experiment with settings, and analyze results.

Evaluating response quality remains your responsibility. The Playground gives you useful tools and measurements, but it does not declare a universally best model for every task.

Next steps

To start using EidoStack:

1. Create your account.
2. Complete workspace setup.
3. Configure your first AI provider.
4. Start your first chat.

Then select an AI model, send your first request, and start exploring the Playground.

For more detail about individual EidoStack features, see:

• Compare model responses.
• System prompts.
• Context management.
• Usage, tokens, and costs.

You can also explore EidoStack features and use cases to see how the Playground helps with coding, prompt testing, exploring new models, and other tasks.

Start with one model, experiment with your own prompts, and explore what AI can do—all in one Playground.