Provider API Keys (BYOK)
Sign up with each supported model provider, generate an API key, and store it on your organization so all inference uses your own credentials.
Open your organization's BYOK panel
Navigate to Organization → Settings → API Keys (BYOK). You need an Owner or Admin role to add or remove keys. Each provider you add is encrypted at rest and only ever used for that organization's chats, automations, dashboards, and embeddings.
Supported providers: OpenAI, Anthropic, Google AI, Groq, Mistral, Cerebras, Azure OpenAI, Cohere.
OpenAI
Create an account at platform.openai.com. Add a payment method under Billing, then open API keys and click Create new secret key. Copy the key (it starts with sk-) — OpenAI will not show it again.
In the BYOK panel, choose OpenAI, paste the key, optionally label it (for example Production), and click Save Key.
Anthropic (Claude)
Sign up at console.anthropic.com. Add credit under Plans & Billing, then open Settings → API Keys and click Create Key. Copy the value (it starts with sk-ant-).
Back in the BYOK panel, choose Anthropic and paste the key.
Google AI (Gemini)
Visit aistudio.google.com and sign in with a Google account. Open Get API key in the left rail (or go directly to aistudio.google.com/app/apikey), create a key in a Google Cloud project, and copy it.
For paid quota and SLA, enable billing on the linked GCP project. In the BYOK panel, choose Google AI and paste the key.
Groq
Sign up at console.groq.com. Open API Keys in the left nav (or go directly to console.groq.com/keys) and click Create API Key. Copy the value (it starts with gsk_).
Groq is a low-latency inference provider that hosts open-weight models (Llama, Qwen, DeepSeek, Mixtral). In the BYOK panel, choose Groq and paste the key.
Mistral
Sign up at console.mistral.ai. Add a payment method under Workspace → Billing, then open API keys and click Create new key. Copy the value.
In the BYOK panel, choose Mistral and paste the key. Available models include mistral-large-latest, mistral-medium-latest, mistral-small-latest, open-mistral-nemo, and codestral-latest.
Cerebras
Sign up at cloud.cerebras.ai. From the platform dashboard, open API Keys in your organization settings and click Create API Key. Copy the value (it starts with csk-).
Cerebras runs Llama and Qwen models on Wafer-Scale Engines and is a good choice when you need very low latency. In the BYOK panel, choose Cerebras and paste the key.
Verify the keys are active
Each saved key appears in the BYOK list with the provider name, optional label, and the date it was added. Start a new chat in your organization — model selection will automatically include the providers whose keys are present and respect any allowed_models / blocked_models policy you set in Enterprise Controls.
To rotate, click Remove next to the provider and add the replacement key. To temporarily fall back to the platform default, just remove the BYOK entry on shared deployments. On dedicated deployments, BYOK is required — there is no platform fallback.