Simple, transparent pricing
Choose your hardware plan based on the resources your agent needs. Add an AI plan for managed model access, or bring your own keys.
Hardware Plans
Your agent runs on a dedicated virtual machine. Choose the hardware that fits your workload.
Starter
Light workloads and personal projects
1,380 ASTD/mo
- All channels (Telegram, Discord, etc.)
- * Highest burstable vCPU
- Suited for text-only use cases
Pro
More power for complex conversations
1,760 ASTD/mo
- All channels
- Suited for a few skills and mild tasks
Max
Heavy workloads and file processing
3,520 ASTD/mo
- All channels
- Suited for automated sophisticated tasks
Business
Maximum performance for demanding use
7,030 ASTD/mo
- All channels + custom integrations
- Suited for heavy multi-agent workloads
- SLA guarantee
💡 Need multiple agents? Purchase a plan for each one.
Save 15% with yearly activation
Pay once for 12 months and get 15% off the monthly ASTD cost. Available when activating your agent.
Need More?
Unlimited resources, dedicated infrastructure, custom integrations, and white-glove support for enterprise teams.
AI Plans
Choose how you access AI models. Bring your own API keys, or use our managed plans with all Claude models included.
BYOK
Bring Your Own Keys
- ✓ Use your Anthropic or OpenAI API keys
- ✓ Access any model your key supports
- ✓ Pay providers directly
- ✓ No markup on API costs
You manage your own API billing
Pay-as-you-go
Use ASTD balance per API call. No monthly commitment.
- ✓ Deducts from ASTD balance
- ✓ All Claude models (Anthropic + Bedrock)
- ✓ Open-source models via Ollama Cloud*
- ✓ Pay only for what you use
1,000 ASTD/mo
Light usage plan
- ✓ 1,000 ASTD monthly credits
- ✓ All Claude models (Anthropic + Bedrock)
- ✓ Open-source models via Ollama Cloud*
- ✓ Charged at cost from provider
1,400 ASTD/mo
Standard-light usage plan
- ✓ 1,400 ASTD monthly credits
- ✓ All Claude models (Anthropic + Bedrock)
- ✓ Open-source models via Ollama Cloud*
- ✓ Charged at cost from provider
3,000 ASTD/mo
Standard usage plan
- ✓ 3,000 ASTD monthly credits
- ✓ All Claude models (Anthropic + Bedrock)
- ✓ Open-source models via Ollama Cloud*
- ✓ Charged at cost from provider
5,000 ASTD/mo
Heavy usage plan
- ✓ 5,000 ASTD monthly credits
- ✓ All Claude models (Anthropic + Bedrock)
- ✓ Open-source models via Ollama Cloud*
- ✓ Priority API access
10,000 ASTD/mo
Maximum usage for demanding workloads
- ✓ 10,000 ASTD monthly credits
- ✓ All Claude models (Anthropic + Bedrock)
- ✓ Open-source models via Ollama Cloud*
- ✓ Priority API access
Available AI Models
View detailed API model pricing →
Anthropic (Direct)
- • Claude 3.5 Haiku
- • Claude Sonnet 4
- • Claude Opus 4.6
AWS Bedrock
- • Claude Sonnet 4
- • Claude Haiku 4.5
- • Claude Opus 4
- • Amazon Nova Pro / Lite / Micro
- • Llama 4 Maverick
- • Llama 3.3 70B
- • DeepSeek R1
- • Mistral Large 2
- • Gemma 3 (27B / 12B / 4B)
- • MiniMax M2
- • Qwen3 32B
- • GLM 4.7 / Flash
- • Kimi K2.5
- • Nemotron Nano 3 30B
- • GPT-OSS 120B
- • Command R+
Open-Source via Ollama Cloud*
- • DeepSeek V3.2 — Free
- • Qwen 3.5 397B — Free
- • Mistral Large 3 — Free
- • Gemma 3 27B — Free
- • GPT-OSS 120B — Free
- • GLM-5 — Free
- • Kimi K2.5 — Free
- • MiniMax M2.5 — Free
* Ollama Cloud models are provided free while usage quota is available. Usage is pooled across all AgentStead users on a shared Ollama Cloud account. When the shared quota is depleted, open-source models may be temporarily unavailable until the quota resets. Anthropic models (Claude) are always available as long as you have ASTD balance. We may upgrade our Ollama Cloud plan as demand grows.
Frequently Asked Questions
What's the difference between the hardware plans?
Each plan gives you a different amount of compute resources. Starter has 0.25 vCPU and 512 MB RAM — good for simple chatbots. Pro and above have more CPU, RAM, and storage for heavier workloads like document processing.
Can I change plans later?
Yes! Upgrade or downgrade anytime. Upgrades take effect immediately (prorated). Downgrades apply on your next billing cycle.
What's the difference between Pro AI and Max AI?
Both include all Claude models (Haiku, Sonnet, Opus). Max AI just gives you 5× the usage limits. Choose based on how much you'll use your agent.
Why do newer models use more of my limit?
Opus is more capable but costs more to run. Think of your AI plan as a daily budget - Opus messages cost more of that budget than Haiku messages.
What happens if I hit my AI limit?
Your agent switches to a lighter model until limits reset the next day. We never cut off your agent mid-conversation.
Is there a free trial?
We offer a 7-day money-back guarantee. If you're not satisfied, we'll refund your first month.