Pricing & Cost Analysis
Overview
ChatSDK pricing is flexible and scales with your usage. This guide explains pricing tiers, cost factors, and how to optimize expenses.
Pricing Tiers
Free Plan
$0/month
- 1 project
- 50 requests/minute
- 500MB documentation storage
- 100 chat sessions/month
- Community support
- Standard rate limits
Best for: Testing, development, small projects
Starter Plan
$25/month
- 3 projects
- 100 requests/minute
- 2GB documentation storage
- Unlimited chat sessions
- Email support
- Basic analytics
Best for: Small production apps, startups
Growth Plan
$99/month
- 10 projects
- 500 requests/minute
- 10GB documentation storage
- Priority email support
- Advanced analytics
- Custom domain (coming soon)
Best for: Growing applications, teams
Pro Plan
$299/month
- Unlimited projects
- 2000 requests/minute
- 50GB documentation storage
- Priority 24/7 support
- Team collaboration
- Advanced integrations
- Custom branding (coming soon)
Best for: High-traffic apps, enterprises
Enterprise Plan
Custom pricing
- Custom limits
- Dedicated support
- SLA guarantees
- Custom integrations
- On-premise options
- Volume discounts
Contact: sales@chatsdk.com
What's Included
All plans include:
- Unlimited chat conversations
- Context-aware responses
- API access
- Documentation management
- Usage analytics
- Monthly invoices
Cost Components
1. API Requests
Rate limits by plan:
| Plan | Rate Limit | Cost |
|---|---|---|
| Free | 50 req/min | Included |
| Starter | 100 req/min | Included |
| Growth | 500 req/min | Included |
| Pro | 2000 req/min | Included |
| Enterprise | Custom | Custom |
2. Documentation Storage
Storage limits:
| Plan | Storage | Overage Cost |
|---|---|---|
| Free | 500MB | $0.05/GB/month |
| Starter | 2GB | $0.05/GB/month |
| Growth | 10GB | $0.05/GB/month |
| Pro | 50GB | $0.05/GB/month |
| Enterprise | Custom | Custom |
3. AI Inference Costs
Important: AI inference costs are NOT included in monthly plans. You pay separately for actual LLM usage.
Claude API Pricing
Claude Sonnet 4.5 (Recommended for production):
- Input tokens: $3.00 per 1M tokens
- Output tokens: $15.00 per 1M tokens
- Prompt Caching: 90% savings on repeated context
- Batch API: 50% discount on non-realtime requests
Claude Haiku 3 (Budget option):
- Input tokens: $0.25 per 1M tokens
- Output tokens: $1.25 per 1M tokens
- Faster, cheaper, less capable
Example: A typical chat with 2000 tokens (input + output) costs:
- Haiku: $0.002 per chat (~$60/month for 30,000 chats)
- Sonnet: $0.009 per chat (~$270/month for 30,000 chats)
4. Embeddings (Optional)
OpenAI Embeddings (used for document vectorization):
- text-embedding-3-small: $0.02 per 1M tokens
- text-embedding-3-large: $0.13 per 1M tokens
Example: Processing 1000 documents (500K tokens total)
- Using text-embedding-3-small: $0.01 (one-time)
5. Team Members
Included for free on Pro plan. Available on lower plans:
- Starter: First 3 members free, $5/additional member
- Growth: First 5 members free, $5/additional member
6. Support
Free: Community support
Starter+: Email support included
Pro: 24/7 priority support included
Enterprise: Dedicated support included
Cost Scenarios
Scenario 1: Early Development
Monthly costs: $0-50
| Service | Cost |
|---|---|
| ChatSDK Free Plan | $0 |
| Claude API (test) | $0-5 |
| Embeddings (docs) | $0-2 |
| Total | $0-7 |
Usage:
- 100-500 test messages/month
- < 100 documents
- Development only
Scenario 2: Small Production App
Monthly costs: $100-300
| Service | Cost |
|---|---|
| ChatSDK Starter Plan | $25 |
| Claude Sonnet ($3/$15 per M) | $50-150 |
| Embeddings (docs) | $1-5 |
| Infrastructure* | $20-50 |
| Total | $96-230 |
Usage:
- 300-1000 active users
- 10-30 daily chat sessions
- 200-500 documents
- ~2000 tokens per chat
Token calculation:
- 30 chats/day × 30 days = 900 chats
- 900 × 2000 tokens = 1.8M tokens
- Cost: (40% input × 1.8M × $3) + (60% output × 1.8M × $15) = $18-24/month
Scenario 3: Growing App (1K-10K Users)
Monthly costs: $500-1000
| Service | Cost |
|---|---|
| ChatSDK Growth Plan | $99 |
| Claude Sonnet | $200-500 |
| Embeddings | $5-15 |
| Infrastructure | $100-200 |
| Support (if needed) | $0-100 |
| Total | $404-914 |
Usage:
- 5K-10K active users
- 100-200 daily chats
- 1000-2000 documents
- More complex queries
Scenario 4: Large Production (10K+ Users)
Monthly costs: $1500-3000+
| Service | Cost |
|---|---|
| ChatSDK Pro Plan | $299 |
| Claude Sonnet | $1000-2000 |
| Embeddings | $20-50 |
| Infrastructure | $200-500 |
| Database & Storage | $100-200 |
| Monitoring & Support | $100-200 |
| Total | $1719-3249+ |
Usage:
- 50K+ active users
- 500-1000 daily chats
- 5000+ documents
- Enterprise features
Cost Optimization Strategies
1. Intelligent Model Routing
Route requests to appropriate model based on complexity:
if (isSimpleQuery(message)) {
use: 'claude-haiku-3' // $0.25/$1.25 per M
} else {
use: 'claude-sonnet-4.5' // $3/$15 per M
}Potential savings: 40-60% on simple queries
2. Prompt Caching
Cache repeated context (product docs, system prompts):
const response = await anthropic.messages.create({
model: 'claude-sonnet-4.5',
system: [
{
type: "text",
text: productDocs,
cache_control: { type: "ephemeral" } // Cached
}
],
messages: [...]
});Potential savings: 90% on cached tokens (typically 50-70% of total input)
Example: If docs are 10K tokens, 50 chats with cached docs:
- Normal cost: 50 chats × 10K tokens × $3/M = $1.50
- Cached cost: 10K tokens × $3/M × 0.1 = $0.03
- Savings: $1.47 per 50 chats
3. Batch API for Async Operations
Use Batch API for non-realtime requests:
- Document summarization
- Batch analytics
- Scheduled reports
Savings: 50% discount on batch requests
4. Efficient Chunking
Optimize document chunking to reduce embeddings:
- Larger chunks (1000-1500 tokens) = fewer embeddings
- Smart overlap (200 tokens) maintains context
- Remove boilerplate before chunking
Potential savings: 30-40% fewer embeddings
5. Metadata Filtering
Use metadata filters to narrow vector search:
await vectorDB.query({
vector: embedding,
topK: 5,
filter: {
projectId: 'abc123',
docType: 'user-guide', // Narrow scope
}
});Benefits: Faster queries, better relevance, lower costs
6. Response Caching
Cache common questions:
const cache = new Map();
const cacheKey = hash(message + context);
if (cache.has(cacheKey)) {
return cache.get(cacheKey);
}
const response = await generateResponse(message);
cache.set(cacheKey, response);
return response;Potential savings: 10-20% on LLM costs
Example Monthly Budgets
Startup (Pre-Revenue)
Budget: $100/month
ChatSDK Starter: $25
Claude API: $50
Embeddings: $5
Infrastructure: $20
Total: $100Growing Startup ($10K MRR)
Budget: $500/month
ChatSDK Growth: $99
Claude API: $250
Embeddings: $10
Infrastructure: $100
Buffer: $41
Total: $500Scaling Company ($100K MRR)
Budget: $2000/month
ChatSDK Pro: $299
Claude API: $1000
Embeddings: $30
Infrastructure: $400
Support & Monitoring: $271
Total: $2000Enterprise ($1M+ MRR)
Budget: Custom
ChatSDK Enterprise: Custom
Claude API: $5000+
Embeddings: $100+
Infrastructure: $1000+
Support & Custom Features: $2000+
Total: $10K+/monthBilling & Invoicing
Invoice Frequency
- Monthly invoices (first day of month)
- Covers usage from previous month
- Paid via credit card or wire transfer
Invoice Contents
- ChatSDK plan fees
- Additional AI inference costs
- Storage overages
- Support fees (if applicable)
Payment Methods
- Credit Card: Visa, Mastercard, American Express
- Wire Transfer: For Enterprise customers
- Monthly or Annual billing (get 15% discount with annual)
Changing Plans
- Upgrade anytime (prorated charge)
- Downgrade at month end
- No long-term contracts (except Enterprise)
FAQs
Q: Can I negotiate custom pricing?
A: Yes! Email sales@chatsdk.com for volume discounts, especially if you're committing to annual plans.
Q: Are there discounts for non-profits?
A: Contact support@chatsdk.com. We offer 50% discounts for qualified non-profits.
Q: What happens if I exceed my rate limit?
A: Your requests are rate-limited (429 response). No overage charges, but service is restricted. Upgrade your plan to increase the limit.
Q: Can I use my own LLM provider?
A: Currently, Claude API is required. Custom LLM support is on our roadmap.
Q: Is there a free trial?
A: Yes! Free Plan lets you test with full features. No credit card required.
Q: What's included in the "infrastructure" cost?
A: Your own servers/hosting for the React app. ChatSDK backend is included in the plan.
Q: Do you offer credits?
A: Yes! New customers get $5 credits for the first month.
Cost Comparison with Alternatives
ChatSDK vs Building In-House
| Feature | ChatSDK | In-House | Savings |
|---|---|---|---|
| Initial Setup | < 1 hour | 4-8 weeks | 8-16 weeks engineer time |
| Monthly Cost | $25-299 | $5,000-15,000 | $4,700-14,975 |
| Maintenance | Included | Dedicated team | Full engineer |
| Support | 24/7 | Internal only | Faster resolution |
| Security | Enterprise-grade | Custom | Better compliance |
Conclusion: ChatSDK is 200-600x cheaper than building in-house.
Next Steps
- Start with Free Plan - No credit card required
- Explore Pro Features - See what you get
- Review Security - Understand data protection
- Contact Sales - Custom pricing questions
Related Documentation
- Admin UI - Manage your account
- API Key Management - Track usage per key
- Analytics - Monitor costs in real-time