GLBNXT Platform / Analytics

See who uses which model, how fast, and at what risk.

Token use per application, API key and model, response times, and a breakdown by AI Act risk level. Filter it your way, open the request log, and export it when someone asks.

Usage Analytics
Last 7 days
Total tokens4.35Mlast 7 days
Requests12,480last 7 days
Avg. time to first token1.7sacross requests
Most used modelglbnxt/gemma-4:26bby tokens
Tokens and requests per dayAll applications · hover a day for details
Tokens (K)Requests
0K250K500K750K1000KPeak · FriMonTueWedThuFriSatSun
Peak dayFri · 830K
Average per day621K tokens
Requests per day1,783
vs last week▲ +12%

Usage per application

Every token has an owner.

Each application calls models with its own key, so every token lands on a name. Click an application to see its share day by day.

0K250K500K750K1000K620540480710830560610MonTueWedThuFriSatSun
Total this week4,350K tokens
Busiest dayFri · 830K
Largest consumerContract review · 40%
Average per day621K

Usage by AI Act risk level

Know how much of your AI runs at High risk.

The risk level you give each API key carries through to your analytics. Filter the dashboard and the log by level, and see the share of tokens per level at a glance.

40%of tokens this week · Level 2 · High risk
Each square = 1% of all tokens

Top lists

The busiest apps, keys and models, ranked.

Three rankings in one place. Spot the application that suddenly doubled, the test key still running in production, or the model everyone quietly moved to.

  1. 1
    Contract review
    1,750 K tokens
  2. 2
    Claims assistant
    1,520 K tokens
  3. 3
    Intranet search
    1,080 K tokens

Speed per model

Pick the right model for the wait.

Time to first token is what your users feel. Fast models keep a chat snappy; deeper ones are worth a few seconds for research and contract work. See where each model lands.

glbnxt/qwen3-embedding:0.6bEmbedding for search
0.3s
glbnxt/gemma-4:26bSnappy chat and search
0.9s
glbnxt/gpt-oss-120bAssistants and drafting
1.6s
deepseek-v4.1-flashCoding and agents
2.1s
glm-5.3Deep reasoning, worth the wait
2.8s
Time to first token
0s1s2s3s4s

Request log

Every request on record. Never the content.

Time, application, model, tokens in and out, status and duration for each call. Prompts and responses are not stored in the log. Filter it, open a row for the details, and export it as CSV.

Request LogEvery request, with its metadata. No prompts or responses.
Date/timeApplicationModelInputOutputStatusDuration
API keyClaims assistant
Risk levelLevel 3 · Limited risk
Time to first token1.4s
Content storedNone, metadata only

One dashboard, three readers

Finance, IT and compliance read the same numbers.

Each team asks a different question. The answers come from the same data, so nobody keeps a spreadsheet on the side.

“Which team is using what?”
  • Token use per application and per API key, so usage has an owner
  • A fair basis for internal allocation across departments
  • Week-on-week trends before the invoice arrives
40%of this week's tokens went to Contract review

Questions

What admins ask about analytics

Missing something? Ask us in the demo and we’ll show you on a live environment.

Does the request log store our prompts?

No. The log keeps metadata only: time, application, model, token counts, status and duration. Prompts and responses are not stored in it.

Who can see the analytics?

Administrators of your organisation. Usage Analytics sits in the platform's admin area, next to API Keys.

Can we filter by AI Act risk level?

Yes. The risk level set on each API key flows into the analytics, so you can filter the dashboard and the request log by level.

Can we export the data?

The request log exports to CSV, with the filters you have applied: date range, application, model and risk level.

Does it show costs?

Analytics shows token use per application, API key and model, which gives you a solid basis for internal allocation. Ask us about cost reporting for your plan.

How far back can we look?

Pick any date range in the dashboard. Retention follows the agreements made for your environment.

Look at your own numbers, not ours.

Start a free trial and watch the dashboard fill up as your first apps connect. Or book 30 minutes and we’ll walk you through it with your use case.

This website and its contents are the exclusive property of GLBNXT. No part of this site, including text, images, or software, may be copied, reproduced, or distributed without prior written consent from GLBNXT B.V. located at Druivenstraat 5-7, 4816 KB Breda, The Netherlands, registered with the Dutch Chamber of Commerce (KvK) under number 95536779. VAT identification number (VAT ID) NL867171716B01. All rights reserved.

This website and its contents are the exclusive property of GLBNXT. No part of this site, including text, images, or software, may be copied, reproduced, or distributed without prior written consent from GLBNXT B.V. located at Druivenstraat 5-7, 4816 KB Breda, The Netherlands, registered with the Dutch Chamber of Commerce (KvK) under number 95536779. VAT identification number (VAT ID) NL867171716B01. All rights reserved.

This website and its contents are the exclusive property of GLBNXT. No part of this site, including text, images, or software, may be copied, reproduced, or distributed without prior written consent from GLBNXT B.V. located at Druivenstraat 5-7, 4816 KB Breda, The Netherlands, registered with the Dutch Chamber of Commerce (KvK) under number 95536779. VAT identification number (VAT ID) NL867171716B01. All rights reserved.

This website and its contents are the exclusive property of GLBNXT. No part of this site, including text, images, or software, may be copied, reproduced, or distributed without prior written consent from GLBNXT B.V. located at Druivenstraat 5-7, 4816 KB Breda, The Netherlands, registered with the Dutch Chamber of Commerce (KvK) under number 95536779. VAT identification number (VAT ID) NL867171716B01. All rights reserved.