Pricing and billing

Billing

Radius charges for what you use. Model requests and paid web search draw from your organization's credit balance.

• Add credit: on your organization's Billing page. The processing fee is shown before you pay. Any member can add credit.
• Automatic top-up: set a balance threshold and a top-up amount, and Radius adds credit whenever the balance drops below it. A processing fee applies to each top-up.
• Credit codes: redeem them on the Billing page.
• Invoices: find your invoices and payment settings on the Billing page after your first purchase.
• If credit runs out: paid requests stop until you add more.

To cap spending for members or groups, see budgets on the Organizations page.

Models

Every model in this table is available to all organizations and paid from your organization's credits. Prices are in US dollars per million tokens.

• Cost per request: depends on the mix of input, output and cache tokens. Cache read and cache write prices apply only when the selected model supports those features.
• Tiered pricing: some models have price tiers depending on the request's total input tokens and applies to the whole request.
• Virtual models such as balanced are charged at the price of the model that answers each request.

Model availability and prices can change.

Anthropic

Model provider	ID	Model training	Data retention	Input token tier	Input	Output	Cache read	Cache write
	claude-fable-5	No	30 days	—	$10.00	$50.00	$1.00	$12.50
	claude-fable-5-1	No	30 days	—	$10.00	$50.00	$0.25	$12.50
	claude-haiku-4-5	No	30 days	—	$1.00	$5.00	$0.10	$1.25
	claude-haiku-5-5	No	30 days	Up to 100,000	$0.10	$0.50	$0.01	$0.125
				Above 100,000	$0.50	$2.50	$0.05	$0.625
	claude-opus-4-5	No	30 days	—	$5.00	$25.00	$0.50	$6.25
	claude-opus-4-8	No	30 days	—	$5.00	$25.00	$0.50	$6.25
	claude-opus-5	No	30 days	—	$5.00	$25.00	$0.50	$6.25
	claude-opus-5-5	No	30 days	—	$4.00	$20.00	$0.20	$5.00
	claude-sonnet-4-5	No	30 days	—	$3.00	$15.00	$0.30	$3.75
	claude-sonnet-5	No	30 days	—	$2.00	$10.00	$0.20	$2.50
	claude-sonnet-5-5	No	30 days	—	$2.00	$10.00	$0.20	$2.50

DeepSeek

Model provider	ID	Model training	Data retention	Input token tier	Input	Output	Cache read	Cache write
	deepseek-v4.1-flash	No	ZDR	—	$0.30	$1.20	$0.006	$0.00

Moonshot AI

Model provider	ID	Model training	Data retention	Input token tier	Input	Output	Cache read	Cache write
	kimi-k3	No	ZDR	—	$3.00	$15.00	$0.30	$0.00

OpenAI

Model provider	ID	Model training	Data retention	Input token tier	Input	Output	Cache read	Cache write
	gpt-5.3-codex	No	30 days	—	$1.75	$14.00	$0.175	$0.00
	gpt-5.4	No	30 days	Up to 272,000	$2.50	$15.00	$0.25	$0.00
				Above 272,000	$5.00	$22.50	$0.50	$0.00
	gpt-5.4-mini	No	30 days	—	$0.75	$4.50	$0.075	$0.00
	gpt-5.5	No	30 days	Up to 272,000	$5.00	$30.00	$0.50	$0.00
				Above 272,000	$10.00	$45.00	$1.00	$0.00
	gpt-5.6-luna	No	30 days	Up to 272,000	$0.20	$1.20	$0.02	$0.25
				Above 272,000	$0.40	$1.80	$0.04	$0.50
	gpt-5.6-sol	No	30 days	Up to 272,000	$4.00	$20.00	$0.40	$5.00
				Above 272,000	$8.00	$30.00	$0.80	$10.00
	gpt-5.6-terra	No	30 days	Up to 272,000	$2.00	$12.00	$0.20	$2.50
				Above 272,000	$4.00	$18.00	$0.40	$5.00
	gpt-6-luna	No	30 days	Up to 272,000	$0.10	$0.50	$0.01	$0.125
				Above 272,000	$0.20	$0.75	$0.02	$0.25
	gpt-6-sol	No	30 days	Up to 272,000	$2.00	$10.00	$0.20	$2.50
				Above 272,000	$4.00	$15.00	$0.40	$5.00
	gpt-6-astra	No	30 days	Up to 272,000	$10.00	$50.00	$1.00	$12.50
				Above 272,000	$20.00	$75.00	$2.00	$25.00

xAI

Model provider	ID	Model training	Data retention	Input token tier	Input	Output	Cache read	Cache write
	grok-4.7	No	ZDR	Up to 200,000	$2.00	$6.00	$0.50	$0.00
				Above 200,000	$4.00	$12.00	$1.00	$0.00

Z.ai

Model provider	ID	Model training	Data retention	Input token tier	Input	Output	Cache read	Cache write
	glm-5.3	No	ZDR	—	$1.40	$4.40	$0.26	$0.00
	glm-5.3-flash	No	ZDR	—	$0.15	$0.50	$0.03	$0.00

How we handle inference data

Radius handles your inputs and outputs to run your requests and return results. Radius does not use them to train models or persist them after processing a request. Inputs include your prompts, conversation history and attached files. Outputs are the responses generated by the model.

The tags on each model’s page describe how the inference providers we work with handle your inputs and outputs:

• No training: Your inputs and outputs are not used to train models.
• Zero Data Retention (ZDR): We define ZDR as no retention of inputs and outputs beyond processing a request and in-memory caching.

OpenAI and Anthropic: We use their standard API terms, with no training on inputs and outputs and 30-day standard retention. See OpenAI’s policy and Anthropic’s policy for details and exceptions.

Hosting: Radius does not currently guarantee a specific processing location. Our inference providers process requests across multiple regions worldwide.

Web search

Radius offers web search powered by Parallel Web Systems. Web search is off by default. Once it's on, each organization can use free, fast search under Parallel's standard terms or paid search with zero data retention (ZDR):

• Free: Fast search at no cost.
• Fast with ZDR: $0.001 per successful search.
• Advanced with ZDR: $0.005 per successful search.

Paid searches use Radius credits. Organization owners and admins can turn on web search and choose the search option by asking Pi or on the organization's Web search page.