PRICING
Your subscription becomes credits.
Usage is metered in credits; each model's reference multiplier is on the menu.
Pay yearly, save 40%. Credits arrive monthly, and leftovers become permanent.
Free$0
1,500 signup credits, never expire · ClawsGO 1.5 Lithium · 2 parallel tasks · Cloud browser & desktop automation
Plus
$9.9/ month
- 30,000 credits every month
- Unlocks Claude Sonnet 5.5 & GPT 5.6 Terra
- 3 parallel tasks · 20 GB disk · Delegate mode
Pro
Most popular
$29.9/ month
- 120,000 credits every month
- Adds Claude Opus 5.5 & GPT 6 Astra
- 10 parallel tasks · 4-core 8G host · 50 GB disk
- Agent cluster mode · GPT Image 2.5 Sunburst images
Max
$99.9/ month
- 500,000 credits every month
- Team collaboration, unlimited members
- 20 parallel tasks · 8-core 16G host · 100 GB disk
- Priority compute at peak hours
Ultra
$299.9/ month
- 2,000,000 credits every month
- Unlocks Claude Fable 5.1
- 50 parallel tasks · 8-core 16G host · 200 GB disk
- Unlimited team members · Priority compute at peak hours
Your fee becomes credits, and higher tiers get more credits per dollar. Monthly credits are valid for the cycle; on yearly plans, whatever is left becomes permanent and stays yours.
Business & Enterprise
Centralized onboarding · Contracts & bank transfer · Enterprise adds private deployment and SLAs — Business $999/mo · Enterprise from $2,999/mo, 40% off yearly
The strongest models, one unit of measure.
- 1.Your subscription becomes credits: about 3,000 per dollar on Plus, 4,000 on Pro, 5,000 on Max and 6,700 on Ultra.
- 2.Every request is metered per token at each model’s own rates. The menu shows a reference multiplier relative to Silicon: ClawsGO 1.5 Lithium x0.5, ClawsGO 1.5 Silicon x1.0, Claude Sonnet 5.5 x1.5, Claude Opus 5.5 x3.0.
- 3.The menu carries the strongest models from each lab: Claude Fable 5.1, Claude Opus 5.5, GPT 6 Astra.
Compute, execution environments, and storage are included in your plan. You only pay for intelligence.
| Model | Reference multiplier | Plan |
|---|---|---|
| ClawsGO 1.5 Lithium | x0.5 | Free |
| Claude Sonnet 5.5 | x1.5 | Plus |
| Claude Opus 5.5 | x3.0 | Pro |
| GPT 6 Astra | x4.0 | Pro |
| Claude Fable 5.1 | x10.0 | Ultra |
Reference multiplier for every model → See models
Long tasks hit the cache a lot, and cached reads cost a fraction of full price.
FAQ