Bring Your Own Key (BYOK) lets you use your own OpenAI or Anthropic API keys with SalesSheet.ai, giving you unlimited AI usage without platform restrictions. This is perfect for power users, large teams, or anyone who wants direct control over their AI costs and model selection.
This guide walks you through obtaining an API key, adding it to SalesSheet.ai, and understanding the benefits and cost implications of using BYOK.
Before you begin
- Create an account with OpenAI (platform.openai.com) or Anthropic (console.anthropic.com)
- Add a payment method to your OpenAI or Anthropic account
- Understand that BYOK usage is billed directly by OpenAI/Anthropic, not through SalesSheet.ai
Obtain Your API Key
For OpenAI:
- Go to platform.openai.com and sign in to your account
- Navigate to API Keys in the left sidebar
- Click "Create new secret key"
- Give it a descriptive name like "SalesSheet CRM"
- Copy the key immediately (you won't be able to see it again)
- Store it securely - treat it like a password
For Anthropic (Claude):
- Go to console.anthropic.com and sign in
- Click on "API Keys" in the account menu
- Click "Create Key"
- Name it "SalesSheet CRM" or similar
- Copy the key immediately and store it securely
Both providers offer free trial credits for new accounts, giving you a chance to test BYOK without cost. Check your account dashboard for current credit balance.
Add Your API Key to SalesSheet.ai
Once you have your API key, connect it to your SalesSheet.ai account:
- Log into SalesSheet.ai and go to Settings > AI > BYOK Configuration
- Click "Add API Key"
- Select your provider (OpenAI or Anthropic)
- Paste your API key into the secure input field
- Click "Validate Key" to test the connection
SalesSheet.ai will verify the key works and display your account status (active, credit balance if applicable). Once validated, the key is encrypted and stored securely. SalesSheet.ai never stores your key in plain text.
You can add keys from multiple providers and set which one to use as default, or let SalesSheet's smart routing choose the best model for each task.
Configure Model Preferences
With BYOK enabled, you can choose which AI models to use for different tasks:
- Fast Models: GPT-4o-mini or Claude 3.5 Haiku for quick queries, summaries, simple tasks
- Powerful Models: GPT-4o or Claude 3.5 Sonnet for complex analysis, long emails, detailed work
- Smart Routing (Recommended): Let SalesSheet.ai automatically choose the best model based on task complexity
To configure model selection:
- Go to Settings > AI > Model Preferences
- Choose "Smart Routing" for automatic optimization, or
- Select specific models for each task category (email drafting, data queries, summarization, etc.)
- Set cost limits if desired (e.g., "Use fast models when possible to minimize costs")
Smart routing saves money by using faster, cheaper models for simple tasks and reserving powerful models for complex work.
Understand the Benefits
Why use BYOK?
- No usage limits: Unlike platform limits, your usage is limited only by your API provider budget
- Cost transparency: See exactly what you're paying for AI at OpenAI/Anthropic rates
- Model selection: Choose specific models and versions, including beta or new releases
- Data privacy: Your API key means your data goes directly to the AI provider you trust
- Enterprise compliance: Some organizations require BYOK for audit and compliance
- Potential savings: Heavy users often save money with BYOK vs. platform pricing
Cost comparison:
Platform plans include AI usage up to a limit, then charge per additional request. BYOK users pay OpenAI/Anthropic directly, typically $0.01-0.15 per 1000 tokens depending on the model. Most users spend $20-50/month with BYOK for moderate usage.
Monitor Your Usage
Track your BYOK usage and costs:
- In SalesSheet.ai: Go to Settings > AI > Usage Dashboard to see token consumption, requests, and estimated costs
- At the provider: Check platform.openai.com or console.anthropic.com for official billing and detailed usage logs
SalesSheet.ai provides real-time usage estimates, but official billing comes from your API provider. Check both regularly to stay within budget.
Set up budget alerts in Settings > AI > Budget Alerts to get notified when you approach spending thresholds (e.g., notify me at $50 spent this month).
You can also view usage breakdowns by task type to see where your tokens are going: email drafting, data queries, summaries, etc.
Pro Tip
Start with smart routing enabled and monitor your costs for the first month. Most users find they spend far less than expected because the AI uses fast, cheap models for 70-80% of tasks. After a month, review your usage dashboard to see if you can optimize further. You can also set up "eco mode" in Settings > AI > Cost Optimization, which aggressively prefers faster models unless you explicitly request detailed analysis.
What to Expect
When using BYOK, you'll experience:
- Unlimited AI usage within your API provider budget
- Slightly faster response times (direct API calls, no platform queuing)
- Access to the latest model versions as soon as providers release them
- Monthly billing directly from OpenAI or Anthropic, not through SalesSheet
- Usage tracking in real-time with 5-minute refresh intervals
- Ability to switch between providers or remove keys at any time
- No change to SalesSheet.ai platform features or pricing
Troubleshooting
Double-check that you copied the entire key without any extra spaces or characters. Ensure your OpenAI or Anthropic account has a valid payment method attached - keys from accounts without payment methods are often restricted. If you recently created the key, wait 2-3 minutes and try again. Check that your API provider account is in good standing (no billing issues).
This means you've hit your API provider's rate limits, which depend on your account tier. For OpenAI, new accounts have lower limits that increase with usage history. Consider upgrading to a paid tier or spreading requests over time. You can also enable "Rate Limit Protection" in Settings > AI > BYOK Configuration, which automatically slows down requests to stay within limits.
Check Settings > AI > Usage Dashboard to see which tasks are consuming the most tokens. Email drafting and long summaries use more tokens than simple queries. Enable "Eco Mode" to prefer cheaper models. Consider setting model preferences to use fast models for routine tasks. Also verify you don't have unnecessary integrations generating AI requests in the background (check Settings > Integrations).