Your prompts stay yours.
r2hub doesn’t rewrite, shorten or add instructions to your prompts. Your content is sent to the model you select.
AI API access · Pay as you goYour models.
Your models.
A clearer API bill.
Access Claude and OpenAI models through r2hub. Connect the tools you use, add credit and pay for your API usage.
No monthly fee. No provider account required.
Claude CodeCodex CLIYour application
Estimate
See the difference.
Start with your monthly API usage. See the estimated cost and saving, with the comparison basis clearly stated.
Loading current account terms…
Our pricing starts with volume discounts from providers. We pass part of that saving on to you. How r2hub works
Pricing & payment terms
Your account prices and checkout total are shown before you add funds.
Models
Choose your model.
Current new-account prices per million tokens. Model IDs, supported features and reference prices in one place.
Loading the public catalog…
Connect
Keep your workflow.
A separate API balance. The tools you already use. Check model availability and supported features before funding.
- 01
Create your account
Check current rates and create an API key in your account.
- 02
Connect your tool
Choose a supported model and set the endpoint and API key.
- 03
Add credit. Make a request.
Top up through Stripe. Review usage, costs and balance as you go.
Preview a model-specific configuration
How r2hub worksVolume pricing.
Volume pricing.
Clear principles.
We buy API tokens from providers at volume discounts and pass part of the saving on to you. Our margin comes from the difference between what we pay and what we charge.
How we handle your dataNo storage. No training at r2hub.
We don’t store your prompts or responses, and we don’t use them to train models. We retain the usage and billing metadata needed to run your account.
Your selected model.
No automatic fallback to another model. Review token counts, request costs and your remaining credit in your account.
How is the estimated saving calculated?
The budget calculator applies the current new-account discount to usage valued at r2hub rates before discount. It is not automatically a comparison with a provider bill. Model-specific provider comparisons appear only with a current source, matching model and matching token categories.
Can I keep using my existing tools?
Yes, where the model and protocol are supported. Update your endpoint and API key; some tools also need a provider profile. Claude Code uses the Messages API, and Codex CLI uses Responses. Check supported features and setup guides.
How can r2hub offer lower prices?
We buy API tokens from providers at volume discounts, then offer access below standard retail rates. Part of the difference is your saving; the rest covers our service and margin. Current prices and available provider comparisons are shown in the model catalog.
Do you change my prompts?
No. r2hub doesn’t rewrite or shorten your prompts, or insert its own instructions. We route your content to your selected model. Protocol fields may be normalized and output limits enforced; these technical adjustments don’t rewrite your prompt content. See supported features.
Does r2hub store my content or use it for training?
No. We don’t store your prompts or responses, and r2hub doesn’t use your API content to train models. We retain usage and billing metadata to operate your account. Read our data handling details.
Does this replace a Claude or ChatGPT subscription?
No. This is separately billed API usage, not the features or allowances of a consumer subscription.
How do payment and top-ups work?
Add credit through Stripe, then pay for metered API usage. Your account shows the current minimum top-up and rates before you add funds. Taxes are calculated at checkout. Auto-reload is optional and requires separate authorization.
What happens if a request fails?
A failure can occur after the provider has done work. Requests with uncertain usage may retain a temporary balance hold during reconciliation. Check the request status or contact support with its ID—never your API key or prompt.
Your next step
Bring your workflow.
Create an account, or talk to us about the models and API usage you need.