Skip to content

Overview

Every AI-powered feature in Contox consumes tokens from your plan's monthly allocation. This page provides a detailed breakdown of the approximate cost per feature, so you can plan your usage and choose the right plan.

Token counts are approximate and vary based on project size, code complexity, and content length.

Plan Allocations

PlanMonthly AI TokensCredits Available
Free50,000No
Personal1,000,000Yes (auto-recharge)
Team5,000,000Yes (auto-recharge)
Business20,000,000Yes (auto-recharge)
EnterpriseUnlimitedYes

When your monthly allocation is exhausted, prepaid credits are used automatically. See Credits for details.

Feature Token Costs

Genesis Scan

Full codebase analysis using Gemini 2.0 Flash. Cost scales with project size.

Project SizeFilesEstimated Tokens
Small~50 files50,000 - 100,000
Medium~150 files100,000 - 200,000
Large~300+ files200,000 - 400,000

Additional costs included in each scan:

Sub-stepModelTokens
Base cost (file walking, structure analysis)-2,000 (fixed)
Embedding deduplicationmistral-embed5,000 - 20,000
Cross-layer LLM dedupgemini-2.0-flash-lite7,000 - 12,000

Security Scan

12-phase enterprise security analysis. Phases 1-7 are deterministic (no LLM), phases 8-12 use AI.

ComponentModelTokens
Base cost (7 non-LLM phases: secrets, SCA, SAST, config audit, etc.)-5,000 (fixed)
Attack surface analysis (Phase 8)gemini-2.0-flash5,000 - 15,000
Taint analysis (Phase 9)gemini-2.0-flash5,000 - 15,000
Deep analysis (Phase 10)gemini-2.0-flash10,000 - 30,000
Dedup/feedback (Phase 11)gemini-2.0-flash3,000 - 8,000
Embedding assembly (Phase 12)mistral-embed2,000 - 10,000
Total per scan30,000 - 80,000

Ask AI

Interactive Q&A with your project brain. Uses Gemini 2.0 Flash.

ComponentModelTokens
Embedding search (context retrieval)mistral-embed~500
Answer generationgemini-2.0-flash2,500 - 5,500
Total per question3,000 - 6,000

Enrichment Pipeline (per commit)

When auto-enrich is enabled, each commit triggers a pipeline of processing steps. These run automatically in the background.

StepModelTokensRuns
Initial enrichmentmistral-small5,000 - 13,000Once per session chunk
Deep enrichmentdevstral-small3,000 - 6,000Per extracted item
Embedding generationmistral-embed2,000 - 5,000Per batch (~20 items)
Deduplicationmistral-medium3,000 - 5,000Per batch
Quality validationmistral-medium3,000 - 5,000Per batch (10 items)
Drift checkmistral-small4,000 - 6,000Once per enrichment
Session summarymistral-small2,000 - 4,000Once per session close
Evidence indexermistral-small~1,000Per item
Topic mergermistral-small~2,000When needed
Total per commit~25,000 - 50,000

On-Demand Features

Features triggered manually from the dashboard or API.

FeatureModelTokens per Call
Hygiene agent (quick mode)mistral-medium5,000 - 9,000
Hygiene agent (weekly mode)mistral-medium18,000 - 36,000
Drift checkmistral-small4,000 - 6,000
Impact analysismistral-small4,000 - 6,000
Onboarding guidemistral-small~3,000

Usage Scenarios

Estimated monthly token consumption for typical usage patterns.

Solo Developer

ActivityFrequencyTokens
Genesis scan (small project)2x/month~150,000
Security scan1x/month~50,000
Auto-enrich (commits)~80 commits~2,400,000
Ask AI~20 questions~100,000
Hygiene (quick)4x/month~30,000
Monthly total~2,700,000

Recommended plan: Personal (1M tokens) with a credit pack, or Team for heavy usage.

Small Team (5 developers)

ActivityFrequencyTokens
Genesis scans (3 projects)6x/month~600,000
Security scans3x/month~180,000
Auto-enrich (commits)~400 commits~12,000,000
Ask AI~100 questions~500,000
Hygiene (weekly)4x/month~120,000
Monthly total~13,400,000

Recommended plan: Business (20M tokens) or Team (5M) with credit packs.

Large Team (10+ developers)

ActivityFrequencyTokens
Genesis scans (10 projects)20x/month~3,000,000
Security scans10x/month~600,000
Auto-enrich (commits)~1,500 commits~45,000,000
Ask AI~500 questions~2,500,000
Hygiene + operationsWeekly~500,000
Monthly total~51,600,000

Recommended plan: Enterprise or Business with auto-recharge credits.

Cost Optimization Tips

  1. Disable auto-enrich on low-priority projects - Toggle off in Project Settings > General to stop automatic token consumption on every commit
  2. Use Genesis scans sparingly - Run after major milestones, not daily. Each scan on a large project can consume 200K+ tokens
  3. Batch your questions - Ask AI context retrieval (embeddings) runs once per question, so detailed questions are more efficient than many small ones
  4. Set member limits - Team admins can set per-member token, scan, and question limits to prevent any single user from exhausting the team quota
  5. Monitor usage - Check Dashboard > Settings > Billing for real-time token consumption breakdown by feature

Member Limits

Team and Business plan admins can set per-member limits:

LimitDescriptionDefault
Token limit (monthly)Max AI tokens per member per monthUnlimited (within plan)
Scan limit (monthly)Max Genesis + Security scans per member per monthUnlimited (within plan)
Ask limit (monthly)Max Ask AI questions per member per monthUnlimited (within plan)

Configure these in Dashboard > Team > Member Settings. A value of 0 means unlimited (constrained only by the team's plan quota).