AI Analytics & Insights: Track Spend, Cut Costs, and Optimize Your AI Usage

# AI Analytics & Insights: Track Spend, Cut Costs, and Optimize Your AI Usage

Most AI platforms leave you guessing about what you actually spent. You run a batch of image generations, trigger a few long reasoning tasks, and somewhere along the way your credit balance quietly shrinks. By the time you notice, you have no idea which models or workflows were responsible.

Vincony's AI Analytics hub solves that. It consolidates everything — raw usage charts, spend breakdowns, cost comparisons, AI coaching, and a monthly Wrapped summary — into one tabbed surface. Whether you are on a Free plan watching every credit or a Business subscriber managing a team budget, the same data is at your fingertips. This guide walks through each of the five panels in detail so you can start making informed decisions immediately.

---

How the Insights Hub Works

Navigate to AI Analytics while signed in. You will see five tabs across the top:

• Usage — your personal generation history and daily activity
• Spend — a full breakdown of credits by category, model, and provider
• Savings — a side-by-side comparison of actual spend vs. what smart routing would have cost
• Coach — personalised, rules-based insights with a 0–100 efficiency score
• Wrapped — a shareable monthly AI personality card

Each tab reads directly from your generations and thread_messages records in real time, so the numbers are always current. None of the panels store data separately; they all query the same source of truth.

---

The Five Panels

1. Usage — Your Personal Dashboard

The Usage tab is the default view when you open Insights. A time-range picker at the top right lets you scope everything to the last 7 days, last 30 days, or last 90 days.

Four summary cards sit at the top:

• Credits Used — total credits consumed in the selected period
• Generations — number of model calls made
• Chat Messages — user turns sent across all threads
• Avg/Day — your average daily credit burn rate

Below the cards, a burn-rate banner gives you a plain-English runway estimate: "At this pace, you'll run out in ~N days" when credits are getting low, or a green confirmation that you are on track. This is especially useful on Starter and Pro plans where your monthly credit allocation is fixed.

An area chart shows credits used over time, so you can spot spikes caused by a heavy project day or an expensive model you tried once. Below that, a Usage by Category section uses horizontal progress bars to show the split between text, image, video, audio, and code generations — each with the exact credit count and percentage.

Your Top Models lists the five models you have called most, ranked by use count and decorated with medal icons. A donut Model Usage Distribution chart lets you click individual segments to isolate one model's contribution, which is useful when you want to understand why a particular model dominates your history.

Finally, a Share Your AI Stats card generates a gradient summary card — showing total credits, total generations, and your favourite model — that you can share via native share sheet or copy to clipboard.

2. Spend — Full Budget Breakdown

The Spend tab goes deeper on the financial picture. It is organised into four inner sub-tabs: Spend Breakdown, Top Models, Optimization, and Routing Presets.

Spend Breakdown shows each category (Text Generation, Image Generation, Video Generation, Agent Tools, Audio/Voice) with a coloured progress bar, exact credit total, and percentage of total spend. Your real usage data is used when available; illustrative examples appear if you are just getting started.

Top Models ranks every model you have used by total credits consumed, with provider attribution and a trend badge (Trending up, Trending down, or Stable). This is the fastest way to see if a new model has quietly become your biggest cost driver.

Optimization presents three nudges: - A savings suggestion when a pattern is detected (e.g., using a large model for tasks where a smaller one would perform equally well) - A warning when a category is consuming a disproportionate share of budget - A success note when your routing choices are already improving

Routing Presets offers three one-click configurations — Cheapest Acceptable, Fastest Acceptable, and Best Quality — that apply a routing strategy across your workspace. A direct link to the Smart Router lets you configure more advanced rules.

At the bottom, a Team Budget Visibility card shows workspace-level credit allocation and per-member averages. Spend alerts (triggered at 80% of budget by default) are available on Business plans, making this section the primary tool for team administrators.

3. Savings — Actual vs. Optimal Cost

The Savings tab is where the most actionable cost data lives. It answers one specific question: "How much more did I spend than I needed to?"

At the top, five summary cards display:

• Actual Spend — what you really paid
• Optimal Spend — what the auto-router would have paid, picking the cheapest capable model for each prompt
• Potential Savings — the gap between the two
• Avg Cost / Gen — your real average vs. the optimal average
• Projected / Month — an extrapolation of your current savings opportunity over 30 days

A Monthly Budget Goal widget lets you set a credit limit for the current month. A progress bar tracks your position; the bar turns amber at 80% and red if you exceed the limit. When you approach the threshold, an inline tip suggests switching to Auto mode. A sparkline chart shows up to six months of historical spend so you can see whether your usage is trending up or down. A 30-day spend heatmap of coloured squares (darker = heavier spend) gives an at-a-glance view of your activity rhythm.

The main Spending Comparison bar chart plots actual credits vs. auto-routed credits side by side, with a toggle to overlay the previous period for trend comparisons. A donut By Category chart lets you click a segment to filter all data to that category.

Top Savings Opportunities identifies specific models where switching would save the most credits. Each row shows the model name, how many times you used it, how many credits were wasted vs. the cheaper alternative, and a Try [model] → button that opens Chat with that model pre-selected.

The collapsible Top Expensive Generations table lists your ten costliest individual calls with prompt previews, dates, and — where a cheaper option exists — a Re-run cheaper button that pre-fills Chat with the same prompt on the more economical model.

If you prefer offline analysis, Export downloads a CSV of your per-model breakdown, Copy copies a plain-text summary to your clipboard, and Share PDF exports the entire report as a formatted PDF.

4. Coach — Personalised Efficiency Insights

The Coach tab analyses your generation history and returns a 0–100 Coach Score. The score is computed from multiple signals — model diversity, fail rate, category distribution, and whether you are consistently using models appropriate for each task type. Green indicates strong optimisation (70+), amber means there is room to improve (40–69), and red flags patterns that need attention.

Four stat cards — Total Credits, Avg Cost, Models Used, and Fail Rate — each carry trend badges comparing the current period to the previous one, so you immediately see whether things are improving or getting worse.

Coach Insights are the core of this panel. Up to a dozen rule-triggered observations are grouped into cards and sorted by severity: saving (green), warning (amber), and info (blue). Each card shows the insight, a detail explanation, and a direct action link (for example, "Go to Settings to enable Auto mode"). Long insights expand inline. You can dismiss any card you have acted on; a Reset dismissed button restores them all.

Two charts help you visualise underlying patterns: - A Credit Usage Over Time line chart tracks daily credits and generation count - A Model Distribution donut chart with an interactive legend lets you highlight individual models

A Credits by Category bar chart shows how spending is distributed across text, image, audio, video, and other categories.

The Credit Savings Calculator at the bottom is a forward-looking tool. Drag the slider to set your expected monthly generation volume (presets: Light 50, Regular 200, Power 500), and the calculator instantly shows your current estimated total, what it would be with Auto mode enabled, and the projected monthly saving. Clicking Enable Auto Mode takes you directly to Settings.

When you are done, the Report button downloads a plain-text file and the Export & Share dropdown lets you save to Google Drive, OneDrive, Dropbox, Notion, or a Google Doc, or send directly via email.

5. Wrapped — Your Monthly AI Personality

The Wrapped tab is a lighter, shareable take on your usage data — inspired by the end-of-year stats formats popularised by music streaming apps, but updated monthly.

Eight animated cards build up with a staggered entrance animation:

• Your AI Personality — one of six archetypes assigned based on your usage (The Explorer, The Model Collector, The Loyalist, The Power User, The Visual Creator, or The Code Wizard)
• Total Generations — combined model calls and chat messages
• Credits Used — total spend for the month
• Favourite Model — your most-called model by name
• Models Explored — count of unique models you tried
• Most Active Day — the weekday you generate most heavily

A Share My Wrapped button triggers the native share sheet on mobile or copies a pre-formatted text card to clipboard on desktop. The card includes your personality type, all six stats, and a link back to vincony.com so people can find their own Wrapped.

---

Step-by-Step: Getting the Most from Insights

• Open AI Analytics and check the Usage tab first. Review your burn-rate banner — if it shows fewer than 30 days remaining, move to the Savings tab immediately.
• Set a monthly budget on the Savings tab. Even a rough number (e.g., 500 credits) activates the progress bar and alert logic.
• Review the Spending Comparison chart in Savings. If the optimal bars are consistently shorter than your actual bars, the gap is your savings opportunity.
• Click into Top Savings Opportunities. For each model listed, click Try [cheaper model] → and run a few test prompts. If quality is acceptable, update your default model or enable Auto mode.
• Check the Coach tab once a week. Focus on red and amber insights first — dismissing them after you have acted keeps the list clean and relevant.
• Use the Savings Calculator before upgrading your plan. If the calculator shows you could halve costs with Auto mode, that is worth trying before spending more on credits.
• Share your Wrapped at the end of each month — it is a lightweight way to reflect on how your AI usage is evolving.

---

Tips for Cutting Costs

• Auto mode is the single highest-impact change. The Savings tab measures exactly how much you could save; the number is often 20–50% for users who default to large models.
• Flash and mini models handle most everyday tasks. Simple summarisation, translation, classification, and short-form writing rarely need a flagship model. Switch via the Smart Router or per-request in Chat.
• Video and audio are credit-intensive categories. If the Spend breakdown shows these consuming a disproportionate share, schedule batch jobs during off-peak hours and avoid one-off test renders on premium models.
• Watch the fail rate in Coach. A high fail rate means credits consumed without output — often caused by sending too long a context to a model with a small context window, or using an unstable model. The Coach will flag this pattern.
• Use the 90-day view in Usage to spot seasonal trends. If one month was unusually heavy, identify the project and consider whether it could be completed more efficiently next time.
• Business plan teams should configure per-workspace and per-member budgets via the Team Budget Visibility section. Proactive alerts at 80% prevent end-of-month surprises.

---

Plans and Credit Allocations

All Vincony plans — Free, Starter, Pro, Power, and Business — share the same Insights hub. The data available scales with usage: the more you generate, the richer the Coach analysis and the more meaningful the Savings comparison. See the pricing page for per-plan credit allocations and details on Business team budgets.

The AI Analytics hub requires a signed-in account. Free users can access all five tabs; Business team administrators additionally see workspace-level budget controls and aggregate team spend metrics.

---

If you have not visited Insights recently, open it now and check the burn-rate banner — it takes about 30 seconds and gives you an immediate read on whether your current usage pattern is sustainable. From there, the Savings tab will show you exactly where your biggest optimisation opportunity is.