If you want to use different AI models inside SIMI, the first step is usually obtaining an API key from the AI provider you want to use.
The good news is that you do not need to search the internet for unofficial key sellers, third-party websites, or complicated workarounds.
The major AI providers already provide their own official platforms where you can create accounts, generate API keys, manage usage, view documentation, and handle billing.
This guide brings those official resources together in one place and then shows you what to do with your API key once you have it.
Many AI providers make their models available through APIs. An API allows another application to communicate with the provider's models programmatically.
SIMI can use these provider APIs to create AI agents based on the models you have access to.
This means you can bring models from different AI providers into one SIMI workspace rather than treating every provider as a completely separate environment.
For example, you could have one SIMI agent connected to an OpenAI model, another connected to a Google Gemini model, another using Claude, and another using a different supported provider.
The exact models and availability can change over time, so always check the provider's current documentation before creating an API key.
Below are links to official provider resources. We recommend using the provider's own website rather than purchasing or obtaining API keys from third-party websites.
OpenAI provides API access to its models through the OpenAI developer platform.
The official OpenAI developer quickstart explains how to create an API key, store it securely, make API requests, and add credits for continued API usage.
Official resources
Keep the key private and follow OpenAI's security guidance. OpenAI specifically recommends not exposing API keys in client-side code or sharing them with other people.
Once you have the key, you can use it when configuring the corresponding provider/model inside SIMI.
Google provides Gemini API access through Google AI Studio.
Google's official documentation provides a dedicated API-key guide, and Google AI Studio provides a direct starting point for obtaining a Gemini API key.
Official resources
Google's current documentation also contains important information about API-key security and the transition from older Standard keys to newer authorization keys, so users should follow Google's current instructions rather than relying on older tutorials.
The key can be used when configuring your Gemini-powered agent in SIMI.
Anthropic provides API access to Claude through its developer platform.
Official resources
When creating a Claude API integration, use Anthropic's official documentation for the current authentication and account requirements.
Enter the appropriate provider credentials in SIMI and configure the agent using the model available through your Anthropic account.
DeepSeek provides API access through its developer platform.
DeepSeek's official API documentation states that an API key must be created before using the API and that API requests use Bearer authentication.
Official resources
DeepSeek also documents compatibility approaches for applications and agent tools, making its documentation particularly useful if you want to understand how its API works beyond simply generating a key.
Use the key when creating the appropriate DeepSeek agent inside SIMI.
Mistral provides API access through its Studio platform.
Mistral's documentation explains that users can create API keys through Studio and that API keys can be managed from the API Keys area.
Official resources
Mistral also recommends storing API keys securely and provides key-management and usage controls through its platform.
Configure the corresponding Mistral agent in SIMI using the credentials and model information required by your SIMI setup.
Groq provides API access through GroqCloud.
The official Groq documentation provides a quickstart for creating an API key and making API requests. Groq also provides an official API-key management page.
Official resources
Add the provider credentials to the relevant SIMI agent configuration.
xAI provides API access to its models through the xAI developer platform.
The official xAI quickstart explains that users can create an xAI account, add credits, generate an API key, and then use that key to make API requests.
Official resources
Use the credentials when setting up the corresponding xAI/Grok agent in SIMI.
Cohere provides API access through its developer platform.
Cohere's official documentation references its API-key dashboard and explains the distinction between trial and production API keys.
Official resources
Cohere also provides documentation covering tool use, agents, retrieval, embeddings, and other capabilities, so its developer resources can be useful beyond simply obtaining an API key.
Configure the corresponding provider and model in SIMI.
Getting the API key is only the first part.
The next step is to bring that provider into SIMI.
The general process is:
This is where the API key becomes more interesting.
Obtaining several API keys normally means that you now have access to several separate AI ecosystems.
SIMI provides a place where those different AI providers can be organized as separate agents.
Instead of thinking only:
you can approach a more useful question:
For example, you could configure different agents for different purposes, compare their responses, bring multiple agents into a group conversation, or use different models as part of a broader workflow.
The usefulness of this approach depends on the capabilities, models, limits, pricing, and API availability of the providers you choose.
One of the reasons users explore multiple AI providers is that different models can have different strengths, pricing structures, context capabilities, tool support, response characteristics, and availability.
That means there may not always be a single model that is ideal for every situation.
With SIMI, you can organize supported providers as separate agents and work with them from one environment.
This changes the workflow from:
into a workflow where the configured providers can be represented by different SIMI agents.
Imagine you are researching a complicated subject.
You could configure several agents using different providers.
You might then ask multiple agents the same question and compare their responses.
You could also use SIMI's multi-agent capabilities where appropriate, allowing several agents to contribute to the same conversation or task.
The objective isn't necessarily to declare that one provider is "the best."
Instead, you can make use of the different models you have access to and decide which outputs are most useful for your particular task.
AI platforms change quickly.
Models are replaced.
API requirements change.
Pricing changes.
Authentication systems change.
New models appear.
Older models are deprecated.
That is why this page intentionally directs you to the official provider resources rather than attempting to reproduce every provider's instructions here.
If a provider changes its API-key creation process, its own documentation is the place where the latest instructions will appear.
This page is intended to make that process easier by bringing the relevant official resources together and showing you what to do with the resulting credentials when working with SIMI.
If you are setting up SIMI for the first time, you can use this page as a starting point:
You don't need to search through dozens of unrelated websites to find where to begin.
The provider remains the source of truth for its API.
SIMI is where you can bring supported AI providers together and turn those individual model connections into a broader multi-agent workspace.
Never purchase API keys from random sellers or accept keys posted publicly online.
Do not publish your own API key in:
Treat an API key as a confidential credential and follow the security recommendations provided by the provider.
If you believe your key has been exposed, use the provider's official key-management tools to revoke, rotate, or replace it as appropriate.
Once you have the API key for a supported provider, the next step is to configure the provider inside SIMI and create your AI agent.
Explore SIMI"The API key gets you access to the model. SIMI gives you a place to organize and work with those AI connections."