Pricing

​

Start Free, Scale When Ready

Xpander starts with a free tier designed for hobbyists and startups to launch projects without worrying about costs. Our mission is to make AI infrastructure invisible. Every account includes monthly limits at no cost:

• 100 threads - Conversation sessions with your agents
• 10 files - Document uploads (up to 50MB each)
• 100 function calls - Tool executions and API calls
• 100 interactions - Agent API requests
• 1M input tokens - LLM prompt processing
• 100K output tokens - LLM response generation
• 5 free serverless agents - Event-driven agents that scale on-demand
• 1 free cloud agent container - Full 1 vCPU and 1GB RAM, always-on
• 1 free cloud MCP container - Full 1 vCPU and 1GB RAM, always-on

​

Pay-As-You-Go Pricing

The free tier works without a credit card. When you exceed free tier limits, you’ll need to add a credit card to continue. This activates pay-as-you-go mode and removes usage caps. How it works:

1. Start free - No credit card required, use the free tier indefinitely
2. Reach free tier limit - Your agents pause when you hit monthly limits
3. Add payment - Add a credit card to remove caps and activate pay-as-you-go
4. Automatic billing - Pay only for usage beyond free tier (billed monthly in packs)

Usage beyond free tiers is billed monthly in packs (fixed bundles of resources at set prices).

Resource	Free Tier	Pack Size	Pack Price
Threads	100	1,000	$10
Files (≤50MB)	10	100	$10
Function Calls	100	100,000	$10
Interactions	100	200,000	$10
LLM Input Tokens	1M	1M	$2.50
LLM Output Tokens	100K	1M	$10
Serverless Agents	5	10	$50
Cloud Agents	1	1 container (1vCPU/1GB)	$20
Cloud MCPs	1	1 container	$20
Builder Seats	0	1 seat	$49

​

How Packs Work

Packs are bundles of resources you purchase when you exceed free tier limits. You only buy what you need - if your usage fits within one pack, you only pay for one pack. Example calculations:

• 150 threads used: 100 free + 50 billable = 1 pack needed (pack contains 1,000) = $10
• 440 threads used: 100 free + 340 billable = 1 pack needed (pack contains 1,000) = $10
• 5,000 threads used: 100 free + 4,900 billable = 5 packs needed (4,900 ÷ 1,000 = 4.9, rounded up) = $50

How it works:

1. Your usage accumulates throughout the month
2. Free tier is consumed first
3. Additional usage is covered by automatic pack purchases
4. You’re billed monthly for total packs consumed

​

Pricing Examples

​

Weekly Report Generator (Free)

Agent that pulls data from 3 sources, analyzes trends, and emails a report to 10 people every Monday. What happens weekly:

• 1 scheduled execution = 1 thread
• Fetch from 3 APIs + analyze + send email = 5 function calls
• 1 scheduled trigger = 1 interaction

Monthly usage (4 weeks):

• 4 threads (within 100 free)
• 20 function calls (within 100 free)
• 4 interactions (within 100 free)

Monthly cost: $0 - completely free

---

​

Database Q&A Agent ($10/month)

AI agent that answers questions from your database. 5 teammates ask 1 question daily (5 messages × 30 days = 150 messages/month). Each query performs 2 database lookups. What happens each day:

• 5 messages from different teammates = 5 threads
• Each message triggers 2 database queries = 10 function calls
• 5 API requests to the agent = 5 interactions

Monthly usage:

• 150 threads: 100 free + 50 billable = 1 pack ($10)
• 300 function calls: 100 free + 200 billable (fits in 100K pack, no extra charge)
• 150 interactions: 100 free + 50 billable (fits in 200K pack, no extra charge)

Monthly cost: $10** or **$0.33/day for your whole team

---

​

One-Time Training Data Generator ($10)

Build a structured dataset of 500 JSON records for model training. Agent runs once, processes 500 items with 30 validation actions per item. What happens:

• 1 agent session = 1 thread
• 500 items × 30 validations = 15,000 function calls
• 1 API invocation = 1 interaction
• No history needed (stateless processing)

Total usage:

• 1 thread (within 100 free)
• 15,000 function calls (100 free + 1 pack = $10)
• 1 interaction (within 100 free)

Total cost: $10 one-time to generate your entire training dataset

---

​

Slack Assistant for Small Team ($59/month)

10 employees use a Slack bot for quick lookups (check PTO balance, find docs, create tickets). Average 2 questions per person per day. What happens daily:

• 20 questions = 20 threads
• Each question hits 1 internal API = 20 function calls
• 20 Slack messages = 20 interactions

Monthly usage (22 workdays):

• 440 threads: 100 free + 340 billable = 1 pack ($10)
• 440 function calls: 100 free + 340 billable (fits in 100K pack, no extra charge)
• 440 interactions: 100 free + 340 billable (fits in 200K pack, no extra charge)
• 1 builder seat = $49

Monthly cost: $59** or **$5.90 per employee for unlimited questions

---

​

Document Processor ($50/month)

Process 200 customer contracts monthly, extract key terms, update CRM. Each contract is 1 PDF with 5 data points to extract. What happens per contract:

• 1 document upload = 1 file + 1 thread
• Extract 5 data points + update CRM = 6 function calls
• 1 processing request = 1 interaction

Monthly usage (200 contracts):

• 200 files (10 free + 2 packs = $20)
• 200 threads (100 free + 1 pack = $10)
• 1,200 function calls (100 free + 1 pack = $10)
• 200 interactions (100 free + 1 pack = $10)

Monthly cost: $50** or **$0.25 per contract processed

​

Resource Definitions

Threads - Conversation sessions or task executions with your agents Files - Documents uploaded to agents (max 50MB per file) Function Calls - Tool executions including API calls, database queries, and custom functions Interactions - Individual messages or API requests to your agents LLM Tokens - Computational units processed by language models (separated into input and output) Serverless Agents - Event-driven agents that scale automatically on-demand Cloud Agents - Always-on containerized agents with dedicated compute (1 vCPU, 1GB RAM per container) Cloud MCPs - Always-on Model Context Protocol servers for custom tool integrations Builder Seats - Platform access for team members who develop and manage agents (end users don’t require seats)

​

On-Premise Deployment

Self-host Xpander on your infrastructure with a prepaid usage model designed for enterprise security and compliance requirements. Self-service installation via Helm Chart or Docker starts at $500 minimum with 5 builder seats included.

​

Prepaid Pack Pricing

Resource	Pack Size	Pack Price
Threads	1,000	$100
Files (≤50MB)	100	$10
Function Calls	100,000	$100
Interactions	200,000	$100
LLM Input Tokens	1M	$2.50
LLM Output Tokens	1M	$10
Serverless Agents	10	$100
Cloud Agents	1 container (1vCPU/1GB)	$30
Cloud MCPs	1 container	$30
Builder Seats	1 seat	$49

​

How On-Premise Billing Works

1. Purchase upfront - Buy prepaid usage packs ($500 minimum)
2. Deploy - Install via Helm Chart or self-hosted Docker
3. Consume - Use resources throughout your billing period
4. Renew - Purchase additional packs when needed
5. Overage - Usage beyond prepaid allocation is billed at renewal

​

Budget Caps

Control spending by setting monthly limits in Settings > Billing. Operations are throttled or paused when your budget cap is reached.

​

Enterprise Solutions

For organizations requiring custom SLAs, volume discounts, dedicated infrastructure, or managed on-premise deployments: Talk to the Team

​

Frequently Asked Questions

​

Get Started