Skip to main content

Free Plan Details

The Alactic AGI Free Plan provides an affordable entry point for individuals, small businesses, and development teams to explore document intelligence and AI-powered content extraction. This guide covers all technical specifications, limitations, and use cases.

Plan Overview​

Target Users:

  • Individual developers and researchers
  • Small businesses processing <50 documents/month
  • Development and testing environments
  • Proof-of-concept projects
  • Students and academic researchers
  • Startups validating product ideas

Not Suitable For:

  • Production workloads at scale
  • Batch processing >50 documents/month
  • Applications requiring GPT-4o model
  • Teams needing advanced features

Technical Specifications​

Compute Resources​

Virtual Machine:

  • SKU: Standard_B2s (Azure B-series Burstable)
  • vCPUs: 2
  • RAM: 4 GB
  • Disk: 30 GB SSD (Standard LRS)
  • Network: Basic tier (no accelerated networking)

Performance Characteristics:

  • Baseline CPU: 40% of vCPU
  • Burst Capacity: 200% of vCPU (accumulates credits during idle periods)
  • Best For: Intermittent workloads with idle time
  • Not For: Sustained high-CPU processing

What this means:

  • Good performance for occasional document processing (few times per day)
  • May slow down during continuous processing (50+ documents in a row)
  • CPU credits replenish during idle periods
  • Perfect for development/testing (not 24/7 production)

Storage Limits​

Document Storage:

  • Capacity: 100 MB total
  • File Size Limit: 50 MB per PDF
  • Retention: Unlimited (until manually deleted)

Cosmos DB (Database):

  • Tier: Serverless
  • Request Units: Up to 5,000 RU/s burst
  • Storage: 50 GB max (typical usage: <1 GB)
  • Best For: Infrequent access patterns

Estimated Capacity:

  • ~50 average PDFs (2 MB each)
  • ~100 small PDFs (1 MB each)
  • ~20 large PDFs (5 MB each)

Processing Limits​

Monthly Quotas:

  • PDFs: 20 documents per month
  • URLs: 50 websites per month
  • Total Processing: 70 documents/month combined

Rate Limits:

  • Concurrent Jobs: 2 at once
  • Queue Size: 50 jobs max
  • Batch Size: 10 PDFs or 25 URLs per batch

Processing Speed: With Standard_B2s VM, expect these times:

Document TypeProcessing TimeNotes
Single-page PDF15-25 secondsText extraction + summarization
10-page PDF45-90 secondsPer document
50-page PDF3-5 minutesMay hit CPU limits
URL (article)10-20 secondsFast websites
URL (complex)30-60 secondsHeavy JavaScript sites

Faster on: Documents processed during low-activity periods (VM has CPU burst credits)
Slower on: Back-to-back processing of many documents (CPU credits depleted)

AI Model Access​

Available Models: Alactic GPT-4o mini: Full access Alactic GPT-4o: Not available (upgrade to Pro for access)

Alactic GPT-4o mini specs:

  • Context Window: 128,000 tokens (~96,000 words)
  • Max Output: 16,384 tokens
  • Input Cost: $0.150 per 1M tokens
  • Output Cost: $0.600 per 1M tokens
  • Quality: Excellent for straightforward extraction and analysis
  • Training: 1 billion+ tokens specialized for documents

Model Limitations on Free Plan:

  • Cannot switch to GPT-4o
  • Cannot use other models (Gemini, Claude, etc.)
  • GPT-4o mini is the only option

API Access​

Full API access included:

  • RESTful API for automation
  • Same authentication as dashboard (deployment key)
  • All endpoints available (process, retrieve, search)
  • No API call limits beyond monthly quotas

API Rate Limits:

  • 10 requests per second
  • 1,000 requests per hour
  • Sufficient for most automation scripts

Example API usage:

# Process document via API
curl -X POST https://<your-vm-ip>/api/v1/process \
  -H "X-Deployment-Key: ak-xxxxx" \
  -F "file=@document.pdf" \
  -F "model=gpt-4o-mini" \
  -F "analysis_depth=standard"

Cost Breakdown​

Infrastructure Costs (Fixed)​

Monthly baseline: $72.00

ResourceSKU/TypeCost/MonthCost/Day
Virtual MachineStandard_B2s$51.00$1.70
VM Disk30GB Standard SSD$4.80$0.16
Cosmos DBServerless$8.00$0.27
Storage AccountStandard LRS$5.00$0.17
NetworkingBasic$3.20$0.11
Total Infrastructure$72.00$2.40

These costs are constant regardless of usage (even if you process zero documents).

Variable Costs (Token Usage)​

Model costs added per document processed:

Example: 10-page research paper

  • Input tokens: ~8,000 (document text)
  • Output tokens: ~500 (summary + analysis)
  • Total tokens: 8,500

Cost calculation (GPT-4o mini):

Input:  8,000 × $0.150 / 1,000,000 = $0.0012
Output:   500 × $0.600 / 1,000,000 = $0.0003
Total: $0.0015 per document

Monthly cost scenarios:

Usage LevelDocumentsAvg TokensModel CostTotal Cost/Mo
Minimal5 docs40k tokens$0.01$72.01
Light10 docs80k tokens$0.02$72.02
Moderate20 docs160k tokens$0.03$72.03
At Quota20 PDFs + 50 URLs500k tokens$0.10$72.10

Key Insight: Model costs are negligible compared to infrastructure costs at Free Plan scale.

Total Cost of Ownership​

First Month:

  • Deployment: $0 (no setup fees)
  • Infrastructure: $72
  • Token usage: ~$0.10 (assuming full quota used)
  • Total: $72.10

Ongoing Monthly:

  • $72.10 if using full quota
  • $72.00 if processing zero documents (still pay for infrastructure)

Annual Cost:

  • $72 × 12 = $864 per year
  • Plus minimal token costs

Cost per Document:

  • Infrastructure: $72 / 70 docs = $1.03 per document
  • Tokens: $0.0015 per document
  • Total: ~$1.03 per document

Comparison to alternatives:

  • AWS Textract: $1.50 per 1,000 pages (~$0.015/page)
  • Google Document AI: $1.50 per 1,000 pages
  • Azure Form Recognizer: $1.50 per 1,000 pages
  • Alactic Free Plan: $1.03 per document (all-inclusive)

Feature Availability​

Included Features​

  • PDF upload and processing
  • URL scraping and extraction
  • Text extraction from documents
  • AI-powered summarization
  • Key information extraction
  • Batch processing (up to 10 PDFs at once)
  • Vector storage for semantic search
  • Full API access
  • Dashboard web interface
  • JSON, Markdown, and Plain Text output formats
  • Standard and Deep analysis modes
  • Email notifications
  • Usage statistics tracking
  • Deployment key authentication
  • Community support (GitHub, forums)

Not Included​

  • GPT-4o model access (Pro plan required)
  • Custom domain names
  • IP whitelisting (Pro+ required)
  • Webhook notifications (Enterprise)
  • Multi-region deployment (Enterprise)
  • SLA guarantees (Enterprise)
  • Priority support (Enterprise)
  • Custom model fine-tuning (Enterprise)
  • Scheduled batch processing (Enterprise)
  • Advanced monitoring (Enterprise)

Use Cases​

Good Fit for Free Plan​

1. Individual Research Projects

  • Academic research: Analyze 10-20 papers per month
  • Market research: Scrape competitor websites
  • Personal knowledge management: Extract key points from articles

Example: PhD student analyzing 20 research papers monthly to build literature review. Uses Free Plan to extract summaries, key findings, and citations from each paper.

2. Small Business Document Processing

  • Invoice processing: 20 invoices/month
  • Receipt management
  • Contract review (low volume)
  • Email attachment processing

Example: Freelance consultant processing client invoices. Uploads 5-10 invoices per week, extracts payment details, due dates, and amounts automatically.

3. Development and Testing

  • Building document processing features
  • Proof-of-concept for SaaS applications
  • Testing before production deployment
  • Evaluating Alactic for future use

Example: Startup validating product idea: legal document analysis tool. Uses Free Plan to test extraction quality, response times, and API integration patterns before committing to production deployment.

4. Content Curation

  • Blog post research: Scrape 50 articles/month
  • Newsletter creation: Extract key points from news sources
  • Social media content: Analyze trending topics

Example: Newsletter creator scraping 50 tech news articles monthly, extracting summaries and key quotes, then compiling into weekly newsletter.

5. Educational Projects

  • Student assignments analyzing documents
  • Teaching AI/ML concepts
  • Classroom demonstrations
  • Learning document processing

Not Suitable for Free Plan

1. Production Applications

  • Can't handle user volume
  • No SLA guarantees
  • Limited processing capacity
  • Infrastructure not scalable

Better fit: Pro or Pro+ plan

2. High-Volume Processing

  • >70 documents per month
  • Continuous batch processing
  • Real-time processing requirements

Better fit: Pro+ or Enterprise

3. Complex Analysis Requirements

  • Need GPT-4o for advanced reasoning
  • Multi-step analysis workflows
  • Custom model fine-tuning

Better fit: Pro (GPT-4o) or Enterprise (fine-tuning)

4. Mission-Critical Workloads

  • Require 99.9% uptime SLA
  • Need priority support
  • Cannot tolerate downtime

Better fit: Enterprise plan

Quota Management​

Monitoring Usage​

Check current usage:

  1. Login to dashboard
  2. Go to Settings → Usage Statistics
  3. View:
    • PDFs processed this month: X / 20
    • URLs scraped this month: Y / 50
    • Storage used: Z MB / 100 MB

Email alerts: Enable notifications when reaching quota:

  • Settings → Notifications
  • ☑ Alert at 75% quota (15 PDFs or 37 URLs)
  • ☑ Alert at 90% quota (18 PDFs or 45 URLs)
  • ☑ Alert at 100% quota

What Happens When Quota Exceeded​

If you try to process 21st PDF:

  • Error message: "Monthly PDF quota exceeded (20/20 used)"
  • Document not processed
  • No charges incurred
  • Must wait until next month or upgrade plan

If you exceed storage (100 MB):

  • Error message: "Storage quota exceeded"
  • Cannot upload new documents until you delete old ones
  • Can still process URLs (no storage impact)

Workaround strategies:

  1. Delete processed documents to free storage:

    • Download results first
    • Delete from dashboard
    • Storage freed immediately

  2. Prioritize URLs over PDFs:

    • URLs don't count against storage
    • 50 URL quota vs 20 PDF quota
    • Use URLs when possible

  3. Schedule processing across months:

    • Quota resets on calendar month (1st of month)
    • Process 20 PDFs end of month, 20 more beginning of next month
    • Effective 40 PDFs in 2-day span

  4. Upgrade to Pro or Pro+ plan:

    • Pro: 100 PDFs + 200 URLs per month
    • Pro+: 500 PDFs + 1,000 URLs per month

Performance Optimization​

Maximize Processing Speed​

1. Process During Low-Activity Periods

  • VM accumulates CPU burst credits during idle time
  • Process in morning (more credits available)
  • Avoid back-to-back processing of 20+ documents

2. Use Quick Extract When Possible

  • 3x faster than Standard Analysis
  • Sufficient if you only need text, not insights
  • Saves tokens (lower cost)

3. Batch Small Documents

  • Batch processing shares overhead
  • 10 × 1-page PDFs faster as batch than individually
  • Use Batch Processing tab

4. Optimize Document Sizes

  • Compress large PDFs before upload
  • Remove unnecessary pages
  • Extract text locally for very large documents (100+ pages)

Reduce Costs​

1. Use Standard Analysis by Default

  • Deep Analysis uses 2-3x more tokens
  • Reserve for complex documents only

2. Enable Vector Storage Selectively

  • Only enable if you'll search these documents later
  • Adds processing time and storage
  • Disable for one-off extractions

3. Minimize Output Formats

  • JSON only (don't generate Markdown + Text + JSON)
  • Download results once
  • Clean up old results

Upgrading from Free Plan​

When to Upgrade​

Upgrade to Pro when:

  • Processing >20 PDFs per month consistently
  • Need GPT-4o for better analysis quality
  • Require faster processing (more vCPUs)
  • Storage quota insufficient (>100 MB)
  • Building production application (need reliability)

Upgrade to Pro+ when:

  • Processing >100 documents per month
  • Need IP whitelisting for security
  • Require significant storage (>5 GB)
  • Production workload with moderate scale

Upgrade to Enterprise when:

  • Mission-critical workload requiring SLA
  • Need priority support
  • Custom requirements (fine-tuning, multi-region, etc.)
  • Compliance requirements (SOC 2, HIPAA, etc.)

How to Upgrade​

Important: Cannot upgrade in-place. Must redeploy.

Upgrade steps:

  1. Backup current data:

    # Export all document results
    # Download from dashboard: Bulk Download All

    # Or via API
    curl -H "X-Deployment-Key: ak-xxxxx" \
         https://<vm-ip>/api/v1/export/all > backup.zip

  2. Document configuration:

    • Deployment key (save securely)
    • API integrations (will need updating)
    • Usage patterns (for sizing new plan)

  3. Delete Free Plan deployment:

    • Azure Portal → Resource Groups
    • Delete entire resource group
    • Or keep resource group and delete resources individually

  4. Deploy new plan from marketplace:

    • Follow Marketplace Deployment
    • Select Pro, Pro+, or Enterprise plan
    • Use same region if possible
    • New deployment key will be generated

  5. Restore data and update integrations:

    • Upload backed-up results (if needed)
    • Update API integrations with new key
    • Update team with new dashboard URL

Downtime: Plan for 30-60 minutes

Cost impact: Pro plan costs $147/month (+$75/month vs Free)

Comparison: Free vs Paid Plans​

FeatureFreeProPro+Enterprise
Price/Month$72$147$295$1,500+
PDFs/Month20100500Unlimited
URLs/Month502001,000Unlimited
Storage100 MB5 GB20 GBCustom
VM vCPUs2 (burst)248+
VM RAM4 GB8 GB16 GB32+ GB
GPT-4o AccessXYYY
GPT-4o miniYYYY
API AccessYYYY
IP WhitelistingXXYY
SLAXXX99.9%
SupportCommunityEmailEmail24/7 Priority

Value proposition:

  • Free: Best $/document for low volume (`<70/month)
  • Pro: Best for individuals needing GPT-4o or moderate volume
  • Pro+: Best for small teams and production workloads
  • Enterprise: Best for mission-critical applications

Frequently Asked Questions​

Q: Can I process 20 PDFs + 50 URLs in the same month?
A: Yes! Quotas are separate. You can process up to 70 documents total (20 PDFs + 50 URLs).

Q: What happens to my data when quota is exceeded?
A: Nothing. Previously processed documents remain accessible. You just can't process new ones until next month or upgrade.

Q: Can I share my deployment with team members?
A: Yes, but all usage counts against single quota. Everyone processes from same pool of 20 PDFs and 50 URLs.

Q: Does quota reset on specific date?
A: Yes, quotas reset on the 1st of each month at 00:00 UTC.

Q: Can I pay for additional quota without upgrading?
A: No. Quotas are fixed per plan. To increase quota, upgrade to Pro or Pro+ plan.

Q: Is Free Plan suitable for production?
A: No. Free Plan is for development, testing, and low-volume personal use only. Use Pro+ or Enterprise for production.

Q: How long can I use Free Plan?
A: Indefinitely. No time limit. Pay $72/month as long as you want.

Q: Can I downgrade from paid plan to Free Plan?
A: Yes, but requires redeployment (same as upgrading). Delete paid deployment and redeploy Free Plan.