Provider list prices. A small fee on credits.
Tokens are charged at exactly what OpenAI, Anthropic, Google and DeepSeek publish. You pay a percentage when you add credits, plus a monthly fee only if your team needs shared workspaces.
Pay as you go
$0 / month
5% fee on credit top-ups
- All models at provider list price
- Minimum top-up $5
- Up to 3 API keys with spend caps
- Usage logs kept 30 days
- Email support within 2 business days
Team
$49 / month
3% fee on credit top-ups
- Everything in Pay as you go
- 10 seats, unlimited API keys
- Roles, audit log and BYOK
- Usage logs kept 90 days, CSV export
- Support within 1 business day
Enterprise
Custom
Volume pricing and invoicing
- Everything in Team
- SAML 2.0 SSO and SCIM
- Provider and region rules
- Custom rate limits and log retention
- Response times set by contract
Model prices per 1M tokens
Standard tier list prices published by each provider, passed through unchanged. Prices last checked October 23, 2026; when a provider changes its price, this table and your billing follow.
| Provider | Model ID | Input | Output |
|---|---|---|---|
| OpenAI | openai/gpt-6-astra | $10.00 | $50.00 |
| OpenAI | openai/gpt-6.1-sol | $2.00 | $10.00 |
| OpenAI | openai/gpt-6-luna | $0.10 | $0.50 |
| Anthropic | anthropic/claude-fable-5.1 | $10.00 | $50.00 |
| Anthropic | anthropic/claude-opus-5.5 | $4.00 | $20.00 |
| Anthropic | anthropic/claude-sonnet-5.5 | $2.00 | $10.00 |
| Anthropic | anthropic/claude-haiku-5.5 | $0.10 | $0.50 |
| google/gemini-4-argon | $2.00* | $10.00* | |
| google/gemini-3.8-flash | $0.75* | $3.75* | |
| DeepSeek | deepseek/deepseek-v4-pro | $1.32 | $3.96 |
| DeepSeek | deepseek/deepseek-flash | $0.30 | $1.20 |
- * Google promotional prices: gemini-4-argon moves to $4.00 / $20.00 from December 1, 2026; gemini-3.8-flash moves to $1.50 / $7.50 from January 1, 2027.
- Claude Haiku 5.5 prices apply to prompts up to 100K tokens; longer prompts are $0.50 / $2.50.
- DeepSeek prices shown are peak hours (Mon–Fri 01:00–04:00 and 06:00–10:00 UTC) on cache miss; off-peak is half price.
How discounts pass through
- Batch
- OpenAI and Gemini batch jobs are billed at 50% of the standard price.
- Prompt caching
- Claude cache writes cost 1.25× input (5-minute) or 2× (1-hour); cache hits cost about 0.1× input, 0.05× on Opus 5.5 and Sonnet 5.5.
- Regional processing
- Where a provider charges extra for regional or US-only processing (for example 1.1× on Claude with US inference), the same multiplier applies.
- Your own keys
- On Team and Enterprise, BYOK usage is billed by the provider and is not deducted from credits.
Billing questions
How is a request priced?
Input tokens × input price + output tokens × output price, using the provider's list price for that model. Cached tokens and batch jobs use the provider's cache and batch rates.
Do credits expire?
Credits stay valid for 12 months from the top-up date. Refunds of unused credits follow our Refund Policy.
What do the top-up fees cover?
Payment processing, gateway infrastructure and support. There is no per-request fee and no markup on token prices.
Can I change plans?
Yes. Upgrades apply immediately and are prorated. Downgrades take effect at the end of the billing month.
See the Refund Policy for unused credits, or contact us for volume pricing.