Pricing

Compare Arrow model prices, calculate token charges, and understand prepaid API billing.

API usage is deducted from your organization’s prepaid balance. Arrow 2.0 and Arrow 2 Telos are billed by token usage; Arrow 1-series models have fixed prices per SVG or vectorization.

Token-usage pricing

Prices below are USD per 1 million tokens and apply across supported Responses and SVG operations.

Model	Input	Cached input	Cache write	Output
Arrow 2.0	$4.00	$0.40	$5.00	$20.00
Arrow 2 Telos	$6.00	$0.60	$7.50	$30.00

Arrow 2.0 has the lower token rates. Arrow 2 Telos offers a larger context window for complex SVG creation at 50% higher rates in every token category. Actual costs depend on the tokens each model uses. Compare model capabilities.

Fixed-price SVG models

Arrow 1-series models charge a fixed amount per successful SVG output or vectorization, rather than by token usage.

Model	SVG generation, per SVG	Image vectorization, per call
Arrow 1.1	$0.20 · 20 credits	$0.15 · 15 credits
Arrow 1.1 Max	$0.25 · 25 credits	$0.20 · 20 credits
Arrow 1.0	$0.30 · 30 credits	$0.30 · 30 credits

All prices are USD, using $0.01 per pricing credit. A generation request with multiple outputs is charged for each successful SVG. Custom funding terms are shown in Billing.

How token charges are calculated

For example, a request using 10,000 uncached input tokens and 2,000 output tokens, with no cache reads or writes, costs:

Model	Input charge	Output charge	Total
Arrow 2.0	$0.04	$0.04	$0.08
Arrow 2 Telos	$0.06	$0.06	$0.12

This is a token-count example, not a fixed price per SVG. Each HTTP request in a Responses tool loop has its own usage and settlement, so producing one final SVG can incur several charges.

Cached input and cache-write tokens are charged at their respective rates instead of the ordinary input rate. Output includes reasoning tokens; reasoning is not charged again separately.

How billing works

Prepaid balance: add funds in Billing. Token charges and fixed-price usage draw from the same API balance; one pricing credit represents $0.01.
Purchases and top-ups: the default minimum purchase is $10. Billing shows your account’s purchase limits and custom funding terms. You can enable automatic top-up when your balance drops below a chosen threshold.
Expiration: purchased credits expire 1 year after purchase.
Separate App billing: App subscription credits cannot be spent through the API. See App plans for monthly subscriptions.

For token-priced requests, $1 of balance is temporarily held before the request runs. Once usage is confirmed, the actual charge is settled and unused funds are released. This hold is not an extra fee, a price estimate, or a maximum charge. See Billing and limits for settlement details.

If your available balance cannot cover a request, the API can return 402 Payment Required before running it. See Errors and debugging.

API billing metadata

Read GET /v1/models for current pricing, model access, and supported operations for your organization and API key.

Token-priced models have billing.kind: "token_usage". Use billing.rates: input, cached_input, cache_write, and output. Rates use millicents_per_million_tokens; divide by 100,000 to get USD per million tokens.
Fixed-price models have billing.kind: "fixed_credit". Use pricing_credits.svg_generate per generated SVG and pricing_credits.svg_vectorize per successful vectorization call. With n successful outputs, generation costs n × svg_generate credits.
Legacy model.pricing strings are deprecated placeholders for older clients. Use the fields above instead.

With rates expressed in USD per million tokens:

Request cost = (uncached input × input rate + cached input × cached rate + cache-write input × cache-write rate + output × output rate) ÷ 1,000,000.

Custom pricing

For higher volumes or custom arrangements, contact sales.

Custom funding terms are reflected in the Billing page before checkout.