Credits & Pricing
Understand Ampere credits, plans, BYOK, and top-ups
Ampere uses credits for managed AI usage.
The core rule is simple:
$1 = 1,000 Ampere credits.
Credits are used for chat models and media generation. Different models cost different amounts because provider prices are different.
How credits work
Chat models
Charged from input and output token usage.
Media models
Charged as a fixed cost per image request.
If you use BYOK, model usage goes through your own provider key. BYOK requests do not use Ampere credits.
Managed AI plans
Current managed plans from the Ampere frontend config:
- Free — $0/mo, 5,000 credits, 2 vCPU, 2GB RAM, 20GB storage
- Pro — $39/mo, 20,000 credits, 4 vCPU, 8GB RAM, 40GB storage
- Ultra — $79/mo, 40,000 credits, 8 vCPU, 16GB RAM, 80GB storage
- Unlimited — $299/mo, 40,000 credits plus unlimited Claude model credits, 12 vCPU, 24GB RAM, 120GB storage
- Business — $499/mo, custom credits, 16 vCPU, 32GB RAM, 200GB storage
Plan features can include hosting, web search, cloud backup, auto update, BYOK support, auto top-up, browser automation, and priority support depending on the plan.
BYOK plans
BYOK means “bring your own key.” Ampere hosts OpenClaw, but model billing goes through your provider account.
Current BYOK plan options in the frontend config:
- BYOK Starter — $9.99/mo, 4 vCPU, 4GB RAM, 40GB storage
- BYOK Pro — $19.99/mo, 6 vCPU, 8GB RAM, 80GB storage
- BYOK Ultra — $39.99/mo, 8 vCPU, 16GB RAM, 120GB storage
- BYOK Max — $59.99/mo, 12 vCPU, 24GB RAM, 200GB storage
- BYOK Enterprise — $89.99/mo, 16 vCPU, 30GB RAM, 300GB storage
Team plans
Team plans are priced per seat and support shared usage.
Managed team plans:
- Team Pro — $39/seat, 20,000 credits, 4 vCPU, 8GB RAM, 40GB storage
- Team Ultra — $79/seat, 40,000 credits, 8 vCPU, 16GB RAM, 80GB storage
- Team Unlimited — $299/seat, 40,000 credits plus unlimited Claude model credits, 12 vCPU, 24GB RAM, 120GB storage
Team BYOK plans use the same BYOK tiers as individual BYOK plans.
Media generation credits
Media models use fixed credit costs per request:
- Seedream 4.5 — 50 standard, 50 HD
- GPT Image 1.5 — 43 standard, 166 HD
- Flux 2 Flex — 63 standard, 125 HD
- Nano Banana 2 — 100 standard, 150 HD
- Nano Banana Pro — 188 standard, 281 HD
Simple media tip
Use Standard while testing ideas. Use HD when you need a polished final result.
Choosing a model
Use cheaper models for simple tasks and premium models for important work.
- For everyday chat, use a budget or balanced model.
- For coding, analysis, or careful writing, use a stronger model.
- For image work, test in Standard before spending HD credits.
- If you are unsure, use Ampere Auto and review the result.