Pricing

Understand credits, model usage, compute charges, and automatic recharge.

Ellipsis uses prepaid credits. Model usage and compute are paid from your balance.

What is charged

Models paid byTokensCPU and memory
Ellipsis creditsThe model's rate, belowThe compute rate, below

Compute costs $0.0504 per CPU core per hour and $0.0162 per GiB of memory per hour, billed per second for the resources allocated while an environment runs, including while it waits for your next message. A paused environment is not billed. Compute is charged to your balance when the environment pauses or stops.

Model pricing

Prices in USD per 1 million tokens, for personal and organization accounts.

ModelInputCached inputOutput
Claude Fable 5.2$8.00$0.20$40.00
Claude Fable 5.1$10.00$0.25$50.00
Claude Opus 5.5$4.00$0.20$20.00
Claude Sonnet 5.5$2.00$0.10$10.00
Claude Haiku 5.5$0.10$0.01$0.50
GPT-6.1 Sol$2.00$0.10$10.00
GPT-6 Astra$10.00$1.00$50.00
GPT-6 Sol$2.00$0.20$10.00
GPT-6 Luna$0.10$0.01$0.50
Gemini 4 Argon$2.00$0.10$10.00
Gemini 3.8 Flash$0.75$0.075$3.75

Cache writes cost 1.25× input for Claude (5-minute) and GPT, or 2× input for Claude (1-hour). Gemini cache writes cost the input price. A GPT-6 or GPT-6.1 request with more than 272K input tokens, cache reads and writes included, costs 2× input and 1.5× output for the whole request. A Claude Haiku 5.5 request with more than 100K input tokens, cache reads and writes included, costs 5× for the whole request. Gemini 4 Argon prices double at 08:00 UTC on December 1, 2026, and Gemini 3.8 Flash prices double at 08:00 UTC on January 1, 2027, for requests sent from then on.

Buy credits

Use Billing > Credits to purchase credits. A 10% platform fee applies to credit purchases before discounts, not to individual sessions.

Credits do not expire.

Automatic recharge

Configure a balance threshold and recharge amount under Billing > Credits. A saved payment method alone does not enable automatic recharge.

Paid requests stop when the available balance or an applicable spend limit is exhausted.

Read usage

Billing > Usage shows model and compute usage separately, and each session shows its own cost.

The usage API groups usage by UTC calendar month.

On this page

Schedule a demo