Pricing Calculator
Estimate your total monthly cost. Platform fee + LLM API costs.
Platform Plans
Starter
$15/mo
Up to 5 agents
- 1-click deploy
- Persistent memory
- Auto-scaling
- Community support
- HTTPS & custom domains
Pro
$25/mo
Unlimited agents
- Unlimited agents
- Analytics dashboard
- Priority support (<4hr)
- Multi-region deploy
- Team collaboration
- API gateway
Scale
$45/mo
Unlimited agents
- Unlimited agents
- Dedicated support (<1hr)
- VPC peering
- SLA 99.99%
- Custom integrations
- Advanced security
LLM API Cost Estimator
LLM API costs are billed directly by the provider. Here are approximate monthly costs based on volume.
| Model | Price | 10K req/mo | 100K req/mo | 1M req/mo |
|---|---|---|---|---|
| GPT-5.6 Sol | $5 / 1M input | $50.00 | $500.00 | $5000 |
| GPT-5.6 Luna | $0.20 / 1M input | $2.00 | $20.00 | $200 |
| Claude Fable 5.1 | $10 / 1M input | $100.00 | $1000.00 | $10000 |
| Claude Opus 5 | $5 / 1M input | $50.00 | $500.00 | $5000 |
| Claude Sonnet 5 | $2 / 1M input | $20.00 | $200.00 | $2000 |
| Claude Haiku 4.5 | $1 / 1M input | $10.00 | $100.00 | $1000 |
| Gemini 3.1 Pro | $2 / 1M input | $20.00 | $200.00 | $2000 |
| Gemini Flash-Lite | $0.30 / 1M input | $3.00 | $30.00 | $300 |
| Mistral Large 3 | $0.50 / 1M input | $5.00 | $50.00 | $500 |
| Mistral Small 3.1 | $0.10 / 1M input | $1.00 | $10.00 | $100 |
Estimates based on ~1K input tokens per request. Actual costs vary by prompt length and output size.
Example Scenarios
Small team — 3 agents, 5K requests/day
Platform: $15/mo (Starter) LLM (GPT-5.6 Luna): ~$30/mo Total: ~$45/mo
Growing team — 10 agents, 20K requests/day
Platform: $25/mo (Pro) LLM (Claude Sonnet 5): ~$1,200/mo Total: ~$1,225/mo
Enterprise — 50+ agents, 100K requests/day
Platform: $45/mo (Scale) LLM (Mistral Small 3.1): ~$300/mo Total: ~$345/mo
4 Ways to Reduce LLM Costs
1. Use model routing — Send simple tasks to cheaper models (GPT-5.6 Luna, Haiku 4.5, Mistral Small 3.1) and complex tasks to premium models.
2. Cache frequent queries — HostAgentes caches identical requests to avoid duplicate LLM calls.
3. Optimize prompts — Shorter prompts cost less. Trim unnecessary context from system prompts.
4. Monitor usage — The dashboard shows cost per agent so you can identify and optimize expensive agents.
Pricing Calculator FAQ
How much does it cost to host an AI agent?
Platform plans start at $15/mo (Paperclip Starter, up to 5 agents). LLM API usage is billed separately by the provider on your own key: at 5K requests/day (~150M tokens/mo), a cheap model like GPT-5.6 Luna adds roughly $30/mo. Use the tables below to size your setup.
Does the calculator include LLM API costs?
Yes — it splits the two cost parts. The HostAgentes plan covers hosting, persistent memory, auto-scaling, and monitoring; the LLM table estimates the provider-billed part based on ~1K input tokens per request. Actual costs vary with prompt length and output size.
Which plan should I pick?
Starter ($15/mo) covers up to 5 agents and fits most first deployments. Pro ($25/mo) adds unlimited agents, multi-region deploys, and priority support. Scale ($45/mo) adds VPC peering, a 99.99% SLA, and dedicated support for production fleets.
Are these LLM prices up to date?
The models and prices in the table (GPT-5.6, Claude Fable 5.1 / Opus 5 / Sonnet 5 / Haiku 4.5, Gemini 3.1 Pro / Flash-Lite, Mistral Large 3 / Small 3.1) were verified against the providers' public pricing pages in September 2026. Providers reprice periodically, so confirm on the provider's pricing page before committing.
What is the cheapest way to run AI agents?
For a single lightweight agent, see our guide to the best AI agent hosting under $5/mo — OpenClaw BYOK plans start at $3.99/mo. For multi-agent production workloads, Paperclip Starter at $15/mo usually beats DIY VPS setups once you count DevOps time.
Ready to start? Pick a plan and deploy your first agent in minutes.
View Plans