Pro
7-day trialFlexible workspace access with recurring platform billing. Includes 10 app-funded searches during the trial and 200 each month after upgrading.
No card required to start. Cancel anytime.
Connect the providers you already use, switch models without starting over, and keep control of your keys and provider costs.
Current model
Claude Sonnet 5
Claude
continuing the same conversationBuilding on that: pair the launch story with one concrete before-and-after example.
Choose monthly access or pay once for the workspace. Connected AI providers continue to bill their usage directly at their own rates.
Compare ChatGPT Plus with API and BYOK costsFlexible workspace access with recurring platform billing. Includes 10 app-funded searches during the trial and 200 each month after upgrading.
No card required to start. Cancel anytime.
Pay once for workspace access with no recurring platform fee. Use your own LinkUp or Exa key for search.
Provider and search usage billed separately.
Included with both plans
Choose a model and approximate your monthly usage. We'll separate the workspace fee from the provider usage you would pay directly.
Estimated Pro monthly total $62.58. Estimated Lifetime ongoing monthly usage $58.58.
Your usage
Approximate a typical month
Anthropic: $10/M input · $50/M output · $1/M cache read · $12.5/M cache write
Estimated cost
Claude Fable 5
300 messages · 250 searches
Pro monthly total
$62.58
$4 workspace fee plus estimated usage
Lifetime ongoing usage
$58.58
Usage only after the $39 purchase
Provider usage
$57.08
10K input / 2K output / 500 reasoning tokens per message
Search usage
$1.50
250 LinkUp calls at $0.006
Lifetime first month: $97.58Break-even: month 10
Estimates cover contexts up to 200K tokens, use published input/output rates, assume 40% of input tokens hit the prompt cache (5% cache writes), and bill reasoning tokens at the output rate. Provider discounts, taxes, and currency conversion can change the actual bill. Search rates are for standard sourced-answer style calls with up to ten results.
Choose hosted encrypted storage or local browser encryption. Hosted keys are decrypted server-side only when needed; local keys cross TLS for a request and are used in memory.
Choose hosted encrypted storage or local browser encryption.
Keys are used when making requests to the connected provider.
Manage or remove provider connections from your workspace.
Direct provider billing. Your connected provider bills your API usage directly.
Add a supported provider key, then choose from its available models inside one workspace.
Also supports OpenRouter and custom OpenAI-compatible endpoints.
Supported now
GPT models
View modelsClaude models
View modelsGemini models
View modelsGrok models
View modelsPro $4/month or Lifetime $39, plus provider usage.
Generate an API key in your provider account.
Add the key securely to your workspace.
Select a supported model and begin.
Understand how keys are handled, who bills your AI usage, and what happens when you connect or remove a provider.
Connect a supported provider and use its models from one workspace.
Pro $4/month or Lifetime $39Provider usage billed separately.
Open workspaceGuides and comparisons
Learn how BYOK works, choose the right model workflow, and connect compatible APIs without paying for multiple overlapping subscriptions.
A practical GPT, Claude, Gemini, and Grok workflow with BYOK setup steps.
Compare subscription pricing with paying providers directly through API usage.
Understand base URLs, model IDs, and how to connect compatible endpoints.
Keep the same conversation when you move between connected AI providers. Your prompt and previous replies stay in context while you choose the model for the next response.
Pro $4/month or Lifetime $39. Provider usage billed separately.
Select a connected provider for the next response. The existing conversation remains unchanged.
Start with one clear audience, demonstrate the workflow, then answer the cost and security objections before asking visitors to begin.
Model for next response