The BYOK plan lets you connect your own API keys from AI providers. You pay your provider directly for AI compute and pay Conduity for the platform.
How It Works
- Sign up and select the BYOK plan ($39/month)
- Go to Settings > API Keys
- Enter your API key from one of the supported providers
- All AI calls from your agents are routed through your key
- You are billed directly by your API provider for token usage
Supported Providers
- Anthropic — Claude Sonnet, Claude Opus, Claude Haiku
- OpenAI — GPT-4o, GPT-4o Mini, and newer models
- Google AI — Gemini 2.0 Flash and newer models
- Any OpenAI-compatible endpoint — Self-hosted models, fine-tuned models, Mistral, Groq, Together, etc.
Why BYOK?
- Cost control — Use your own negotiated rates or startup credits
- Model flexibility — Connect any model, including self-hosted or fine-tuned
- Unlimited messages — No message limits on Conduity's side
- Data routing — Your data goes directly to your API provider under your own terms
What You Still Get from Conduity
The BYOK plan includes everything in the Team plan except custom chat branding (available as an add-on):
- Unlimited agents, channels, integrations, and automations
- Memory and knowledge base
- 5 workspace members (additional seats at $12/month)
- Audit logs
- Conversation history
Key Management
- Keys are encrypted at rest using AES-256-GCM
- Keys are never logged or exposed in the UI after entry
- You can rotate keys at any time from Settings
- If your key is invalid or has insufficient credits, you'll see a clear error message
Estimating Your API Costs
A typical Conduity message uses approximately:
| Component | Tokens |
|---|---|
| System prompt + context | ~2,000 input |
| Agent response | ~400 output |
At Claude Sonnet 4 rates ($3/$15 per million tokens), this is roughly $0.01–0.02 per message. Actual costs vary by model, conversation length, and whether knowledge retrieval is involved.