AI Usage Analysis Dashboard

Budget forecasting for tools in Niteco

Load Cursor Data Loading... Load v0 Data Loading... Load GPT Data Loading...

Processing data...

Total Requests

Per Seat / Month

p10-

avg0

p90-

Per User / Month

p10-

avg0

p90-

Total Tokens

In/Output + Cache Read/Write

Per Seat / Month

p10-

avg0

p90-

Per User / Month

p10-

avg0

p90-

Total Tokens (w/ Cache)

Cache Read + Cache Write

Per Seat / Month

p10-

avg0

p90-

Per User / Month

p10-

avg0

p90-

Total Tokens (w/o Cache)

Input + Output only

Per Seat / Month

p10-

avg0

p90-

Per User / Month

p10-

avg0

p90-

Simulated Cost

$0.00

0 users

Per Seat / Month

p10-

avg$0.00

p90-

Per User / Month

p10-

avg$0.00

p90-

💡 Understanding Simulated Cost

Current Seat-Based Deal

Secured March 2025 (expires March 29, 2026)

$11,520/year

$32 × 30 seats × 12 months
(Annual billing discount)
✓ Unlimited slow requests
$6.40/user/month

Simulated Cost with New Usage-Based Pricing

Based on actual usage + token fees

$0/year

Calculating...

$0/user/month

Annual Savings with Current Deal: $0

Deal expires: April 2026

Usage by Model

Token Distribution

Cost Over Time

Tokens Over Time

Requests Over Time

Cost by Seat

Seat	Requests	Input Tokens	Output Tokens	Cache Read	Cost	% of Total

Cost by Model

Model	Requests	Input Tokens	Output Tokens	Cost	% of Total

Monthly Summary

Month	Requests	Input Tokens	Output Tokens	Cache Read	Cost

2026 Cost Simulation

Projecting costs based on historical averages with 2% monthly growth

What-if: Model Distribution Scenario

Current Distribution (Historical) Based on historical usage patterns What-if A: Balanced Less Sonnet, more Auto & Composer-1 (50/25/25) What-if B: Optimized Minimum Sonnet, even more Auto & Composer-1 (20/40/40)

Base Usage Metrics

Avg Monthly Tokens/User 0

Usage Cost/User/Month $0.00

Usage Cost/User/Hour $0.00

Configuration

Total Cursor Users Number of Cursor licenses needed

Hybrid: Claude Users Power users on Claude Team Premium

Plan Cost Comparison

Cost Component	Cursor Team	Cursor Enterprise	Hybrid Option

🔀 Hybrid Option Strategy

Top 3 highest-consuming Cursor seats (3 seats × 5 users = 15 users) moved to Claude Team Premium
15 users get Claude Team Premium @ $125/user/mo = $1,875/mo
- Unlimited Claude API usage (no token fees)
- Direct API access for automation and custom tools
- NO Cursor access (seats removed)
27 Cursor Enterprise seats @ $40/seat/mo = $1,080/mo
- Remaining 135 users (150 - 15) share 27 seats
- ~5 users per seat on average
- Reduced token fees (top consumers removed)
Cost optimization: High-volume users get unlimited access via Claude Team Premium, eliminating their Cursor token fees while maintaining IDE integration for broader team
Total fixed cost: $1,875 + $1,080 = $2,955/mo ($35,460/year)

Plan Features

Feature / Pricing	Cursor Team	Cursor Enterprise

2026 Claude Simulation

Projecting costs for Claude subscription plans based on historical averages with 2% monthly growth (seats = users)

Note 1: Subscription-based plans with strict no seat-sharing enforcement. Anthropic does not publish exact token limits — values shown are community-estimated.

Note 2: Token & request counts from Cursor may differ from Claude due to different counting methods. The numbers are just for reference.

Team Configuration

Number of Users Team plans limited to 75 users

What-if: Extra Usage Distribution

What-if A: 100% Sonnet 100% Claude Sonnet 4.5 What-if B: 40% Opus, 60% Sonnet 40% Opus 4.5, 60% Sonnet 4.5 What-if C: 100% Opus 100% Claude Opus 4.5

Plan Cost & Capacity Analysis

Plan	Seat Cost ?Base monthly subscription per user	Incld Capacity ?Tokens included per user/month (Mtok)	vs Avg ?Plan vs your avg: - P75: - P90: -	Extra Monthly Needed ?Overage beyond included capacity Cache discount: - Cost Formula: -	Per User Monthly Cost ?Seat + overage @ API rates	Total Monthly Cost (- Users) ?Per user × - users	Annual Cost (2% Growth) ?Seat cost (fixed) × 12 months + overage/usage with 2% monthly growth rate. Only variable usage costs grow, not seat costs.
Team Standard Annual billing	$25/user/mo Annual billing	$180 credits/mo ⚠️ Subject to limits	- - \| -	-	-	-	-
Team Premium Annual billing	$125/user/mo Annual billing	$900 credits/mo ⚠️ Subject to limits	- - \| -	-	-	-	-
Enterprise Pay-as-you-go shared pool	$20/user/mo	-	N/A	-	-	-	-

* Hover over column headers (?) for detailed calculation explanations

⚠️ Important: Included Credits Subject to Rate Limits

Team plans include monthly credits ($180 for Standard, $900 for Premium), but usage is constrained by:
• 5-hour session limit - After 5 hours of continuous use, you must wait before resuming
• Weekly usage caps - Credits cannot be consumed evenly throughout the month
• Request rate limits - Maximum requests per minute/hour

These restrictions mean you may not fully utilize included credits even if your average monthly usage "fits" within the plan capacity. For unrestricted access, consider Enterprise (pay-as-you-go at API rates with no session or weekly limits).

Data Training & Retention Policy

Feature	Team	Enterprise
Training Policy
Data Used for Training? Prompts and outputs used for model training	NO - Commercial Terms Contractually guaranteed	NO - Commercial Terms Contractually guaranteed
Opt-Out Available? Need to request training exclusion	N/A - protected by default No action needed	N/A - protected by default No action needed
Retention Policy
Default Retention How long conversation data is kept	30 days Fixed, cannot change	Indefinite (configurable) Min 30 days or indefinite
Custom Retention Set different periods per use case	NOT AVAILABLE	YES - minimum 30 days E.g., 90d for R&D, 180d for support
Zero Data Retention (ZDR) Immediate deletion upon completion	NOT AVAILABLE	YES - via security addendum Critical for GDPR/CCPA compliance
User Deletion When user is removed from system	Within 30 days Full retention period applies	Immediate with ZDR Instant data purge available

For Niteco: Both Team and Enterprise plans are protected under Commercial Terms of Service. Anthropic contractually guarantees your data will NOT be used for model training.

📘 Policy Details & Clarifications

Commercial Terms Protection: "NO - Commercial Terms" means Anthropic contractually guarantees your prompts and outputs will NOT be used to train their models. This protection is automatic and built-in for both plans—no opt-out needed.

Default Retention (Team): All conversation data is retained for exactly 30 days, then automatically deleted. This period is fixed and cannot be changed.

Custom Retention (Enterprise): Allows different retention periods for different use cases or departments. Examples: 90 days for R&D projects, 180 days for production support, indefinite for audit trails.

Zero Data Retention (ZDR): Available via Enterprise security addendum. When enabled, conversation data is deleted immediately upon completion rather than waiting 30 days. Critical for client data processing, regulatory compliance (GDPR/CCPA), and contractual obligations requiring immediate data deletion.

User Deletion: Team plan retains data for the full 30-day period after user removal. Enterprise with ZDR can purge user data immediately when the user is removed from the system.

2026 GitHub Copilot Simulation

Projecting costs for GitHub Copilot plans based on historical averages with 2% monthly growth

Note: Token & request counts from Cursor may differ from GitHub Copilot due to different counting methods. The numbers are just for reference.

Team Configuration

Number of Developers

Plan Cost & Capacity Analysis

Plan	Seat Cost ?Base monthly subscription per developer	Incld Capacity ?Requests included per user/month	vs Avg ?Plan vs your avg: - req/mo P75: - req/mo P90: - req/mo	Extra Monthly Needed ?Additional requests beyond included capacity (avg - included)	Total Monthly (- Users) ?Per developer × - developers	Annual Cost ?Monthly cost × 12 months (fixed seat cost)
Business Organizations	$19/user/mo	300 req/mo	- - \| -	-	-	-
Enterprise Copilot + GitHub Enterprise Cloud	$60/user/mo	1,000 req/mo	- - \| -	-	-	-

Note: Annual cost = monthly cost × 12 (fixed seat pricing, no usage-based overage).

Enterprise pricing: $39 Copilot Enterprise + $21 GitHub Enterprise Cloud = $60/user/mo total.

Feature Comparison by Plan

Feature	Business	Enterprise
Enterprise & Governance
Organization license management	✓	✓
Policy management	✓	✓
Public code filter	✓	✓
Fine-tuned custom models	✗	✓
IP indemnity	✗	✓
SSO & SCIM	✗	✓
Audit logs	✗	✓
Data & Privacy
Training on your code	✓ No	✓ No

2025 Usage Summary

Total Requests

Monthly Average

By Model

Total Cost

$0.00

0 users

Monthly Average

$0.00

Avg Cost / Request

$0.00

By Model

Model Distribution

Usage by Model

Cost Distribution

Cost by Project

Enable grouping to combine related projects (e.g., JOST - Trang, JOST-phase2a → JOST). Grouped projects show a count (5) and you can hover to see all included names.

Project	Requests	Total Cost	% of Total

Cost and Requests Over Time

2026 Cost Projection

Model Configuration

Jost Distribution (most intensive)

Based on historical usage

What-if: Project Configuration

Estimate costs by specifying expected number of active projects per tier

Light Projects

10-50 req/mo

Medium Projects

50-150 req/mo

Heavy Projects

150-400 req/mo

Intensive Projects

400-1000 req/mo

Cost Calculation Breakdown

AVG COST / REQUEST

$0.00

= Jost Total Cost / Jost Requests

PROJECTED MONTHLY REQUESTS

0 - 0

= Σ project tier requests

ESTIMATED MONTHLY

$0 - $0

= Avg Cost × Projected Requests

2026 ANNUAL PROJECTION

$0 - $0

= Σ Monthly × (1.02)^n, n=0..11

Niteco GPT Usage Summary

Total Requests

Average Monthly

total0

Per User / Month

p10-

avg0

p90-

Total Tokens

Input + Output

Average Monthly

total0

Per User / Month

p10-

avg0

p90-

Estimated Cost

$0.00

0 users/keys

Average Monthly

total$0.00

Avg Cost / Request

avg$0.00

Per User / Month

p10-

avg$0.00

p90-

Model Distribution

Token Distribution

Usage by User / API Key

User / API Key	Requests	Input Tokens	Output Tokens	Total Tokens	Est. Cost	% of Total

Usage by Model

Model	Requests	Input Tokens	Output Tokens	Est. Cost	% of Total

Cost Over Time

Tokens Over Time

Requests Over Time

2026 Cost Projection

What-if: 2026 User Distribution

Based on 2025 historical data. Ultra users feature 5x request rate with advanced tools and premium models.

Ultra Users

p90 × 5: $0.00/user/month

Heavy Users

p90: $0.00/user/month

Standard Users

p60: $0.00/user/month

                                Total Users
                                310
                                2025: 0 active
                            

What-if: 2026 API Key Distribution

Based on 2025 historical data. Configure expected number of API keys for each tier.

Heavy Projects

p90: $0.00/key/month

Standard Projects

p60: $0.00/key/month

                                Total Projects
                                7
                                2025: 0 active
                            

Cost Calculation Breakdown

User Costs

Ultra Users

$0.00

30 users

30 users × $0/mo

Heavy Users

$0.00

30 users

30 users × $0/mo

Standard Users

$0.00

250 users

250 users × $0/mo

Users Subtotal

$0.00

310 users

Monthly estimate

Project Costs

Heavy Projects

$0.00

20 keys

20 keys × $0/mo

Standard Projects

$0.00

50 keys

50 keys × $0/mo

Projects Subtotal

$0.00

70 keys

Monthly estimate

Monthly Grand Total

$0.00

310 users + 70 API keys

Annual Grand Total

$0.00

12 months @ 2% monthly growth

AI Coding Tools Comparison: GitHub Copilot vs Cursor vs Claude

Fact-based comparison of three fundamentally different AI coding approaches
Last updated: February 2026

Market Position & Customer Base

Metric	GitHub Copilot	Cursor	Claude (Anthropic)
Primary Customer Segments	• Enterprise, Fortune 100 • Microsoft/GitHub ecosystem • Regulated industries	• Tech startups, scale-ups • Individual power users • Fortune 500 companies	• Enterprise • Mixed technical/non-technical teams • Finance, healthcare, legal sectors
Notable Customers	• Most Fortune 100 • Microsoft internal teams	• Midjourney, Perplexity, OpenAI • Shopify, Stripe, Salesforce, Infosys, EPAM	• Bridgewater, BCG • DoorDash, Sourcegraph, Asana

What Each Tool Actually Is

Category	GitHub Copilot	Cursor	Claude (Anthropic)
Product Type	AI coding assistant plugin for existing IDEs	AI-native IDE (VS Code fork with built-in AI)	General-purpose AI platform with coding as one vertical
Integration Model	Plugin you add to VS Code, Visual Studio, JetBrains, Neovim, etc. Keep your existing workflow.	Replace your current IDE. Move to Cursor. Entire editor designed around AI-first workflows.	Desktop app, Claude Code (terminal), VS Code extension, Browser agent.
Core Functionality	AI suggests code inline as you type. Code completions, chat, PR workflows.	Parallel agents, multi-file refactoring, background task execution, full codebase context.	Agentic coding via terminal And Chat, computer use, document generation (Artifacts), desktop automation.

Exceptional/Differentiating Features

Feature	GitHub Copilot	Cursor	Claude (Anthropic)
Enterprise Features	SSO, SCIM, audit logs, IP indemnity	SSO, audit logs, model management	SSO, SCIM, audit logs, SOC 2 Type II, HIPAA eligible
Agentic Capabilities	Ecosystem Agent • Issue-to-PR automation • Agentic DevOps (security fixes in CI/CD) • Enterprise-constrained autonomy	IDE Orchestrator • Composer Agent Mode (auto file edits + terminal) • Parallel agents	Autonomous Operator • Terminal-based system orchestration • Sub-agent delegation (parallel tasks)
Context & Intelligence	64K tokens standard, file-level focus	Codebase indexing, multi-file context	200K standard, 500K Enterprise
Model Flexibility	Multi-provider	Multi-provider (most broad support)	Claude models primary
Cost Model	Request-based, no share seat, extra @ $0.04/request starting from • Business $19/seat/mo - 300/req/mo • Enterprise $(39+21)/seat/mo - 1000/req/mo	Credit-based, charge @ api rate + fee @ $0.25 per MTok, starting from • Teams $40/seat/mo ($20 individual credit) • Enterprise $60/seat/mo ($20 shared credit)	Subscription-based, no share seat, extra @ api rate, starting from • Team $25-125/seat/mo ($160-$900 worth of credit) • Enterprise 20/seat; (Usage base available)

Limitations & Concerns

Concern	GitHub Copilot	Cursor	Claude (Anthropic)
Key Issues	• Limited context • Weak multi-file refactoring • Less autonomy	• Performance issues (large files, memory freezes)	• Terminal-based • Usage limits controversy

Configuration

User Count 150

Hours per Month 166

Usage Cost/User/Hour -

Token Fee (Base) $0.25 / M tokens

Team Plan Fee $0.25 / M tokens (no discount)

Enterprise Fee (≤$100k) $0.225 / M tokens (10% discount)

Enterprise Fee (>$100k) $0.2125 / M tokens (15% discount)

Model Pricing ($ per million tokens)

Model	Input	Cache Write	Cache Read	Output	Token Fee

Claude API Cache Pricing (for Extra Usage)

This pricing is used when calculating costs for extra tokens beyond included plan capacity. Prices are per million tokens, based on official Claude API pricing.

Model	Cache Read	Cache Write	Regular Input	Output	Cache Benefit
Sonnet 4.5	$0.30	$3.75	$3.00	$15.00	10x cheaper reads
Haiku 4.5	$0.10	$1.25	$1.00	$5.00	10x cheaper reads
Calculated Distribution from Your Data
Input/Output Ratio		99.2% input / 0.8% output
Cache Read (of input)		88.9%
Cache Write (of input)		9.3%
Regular Input (of input)		1.8%

* Cache reads are 10x cheaper than regular input, making high cache hit rates very cost-effective
* Distribution percentages are calculated automatically from your uploaded CSV data

💡 Understanding Simulated Cost

Usage by Model

Token Distribution

Cost Over Time

Tokens Over Time

Requests Over Time

Cost by Seat

Cost by Model

Monthly Summary

2026 Cost Simulation

What-if: Model Distribution Scenario

Base Usage Metrics

Configuration

Plan Cost Comparison

🔀 Hybrid Option Strategy

Plan Features

2026 Claude Simulation

Team Configuration

Plan Cost & Capacity Analysis

Data Training & Retention Policy

2026 GitHub Copilot Simulation

Team Configuration

Plan Cost & Capacity Analysis

Feature Comparison by Plan

2025 Usage Summary

Model Distribution

Usage by Model

Cost Distribution

Cost by Project

Cost and Requests Over Time

2026 Cost Projection

Model Configuration

What-if: Project Configuration

Cost Calculation Breakdown

Niteco GPT Usage Summary

Model Distribution

Token Distribution

Usage by User / API Key

Usage by Model

Cost Over Time

Tokens Over Time

Requests Over Time

2026 Cost Projection

What-if: 2026 User Distribution

What-if: 2026 API Key Distribution

Cost Calculation Breakdown

User Costs

Project Costs

AI Coding Tools Comparison: GitHub Copilot vs Cursor vs Claude

Market Position & Customer Base

What Each Tool Actually Is

Exceptional/Differentiating Features

Limitations & Concerns

Configuration

Model Pricing ($ per million tokens)

Claude API Cache Pricing (for Extra Usage)

Usage Categories

Billable (Included)

Free

Excluded from Analysis