NemulAI is a billing audit for multi-tenant GPU providers. A lightweight, read-only agent attributes real cost — compute + energy — to every job and rolls it up by tenant, so you can see which accounts you're billing below cost and exactly what to charge instead.
Two real MI300X configs of the same Qwen2.5-7B-Instruct LoRA fine-tune (2026-05-07): same 245,760 tokens, but bs=2 ran 23% more energy-efficient — 0.36 vs 0.46 J/token. See the full sample report →
Built for neoclouds, GPU resellers & internal platform teams that bill tenants for shared GPU — anyone who needs to defend margin and prove their true cost of goods.
Shared nodes, fractional GPUs (MIG), and idle time make per-tenant cost a guess. You bill from a rate card and hope it clears cost — with no way to prove what any account actually consumed.
Aggregate cloud bills and scheduler logs that never map GPU-seconds to dollars per tenant. You discover the underpriced accounts in the quarterly margin review — if at all.
A read-only agent ties every GPU-second and its real energy cost to a job → tenant, flags the accounts billed below cost, and tells you the break-even rate to charge.
How it works
One tenant, one number, five steps. Scroll to follow acme through the whole pipeline — then flip the switch yourself.
This is the real energy acme's jobs burned — read from the GPU's own counters, not guessed from an invoice.
Integrated over time, watts become the joules acme actually consumed.
↓ scroll to walk acme through the pipeline
Flip the switch. On energy alone, every account looks healthy. On true cost, two of them are underwater.
Looks fine, doesn't it? That's the trap.
Hover any card to see how deep it reads your data. Green means a healthy margin; red means money is leaking or a customer is underpriced.
Find who is underpriced.
3 files in → customer margin vs. GPU-seconds consumed, with the accounts billed below cost flagged first.
Tie every dollar to a job.
Per run → which model, on which GPU, for how long, at what energy. Nothing lands in the bill unattributed.
Charge the right amount.
Per customer → measured break-even $/GPU-hr and a suggested list price to hit your target margin.
Catch the idle burn.
Per machine → idle %, dollars leaking, and the run that should have been scheduled there instead.
Budget alerts via email and webhook. Carbon & EU AI Act reporting available for teams that need it.
of GPU-hours typically sit idle or underused on the fleets we've measured — capacity you paid for but never billed a tenant. That's the margin NemulAI recovers first.
Set your blended cost and fleet size. NemulAI replaces this estimate with real, measured cost per tenant — and shows exactly which accounts and machines the leak comes from.
Estimate only — NemulAI prices your real, measured energy draw at the rate you set above.
Lightweight, read-only, and reversible. You only need telemetry — not permission to touch production workloads.
pip install the read-only agent on your boxes — systemd, Docker, or a K8s DaemonSet. No workload changes.
Set two env vars so cost rolls up by tenant and model. Scheduler metadata (SLURM, K8s, MIG slices) is picked up automatically.
The agent attributes every job's real cost to a tenant and flags accounts running below cost as it happens.
You get per-tenant cost vs. what you billed, the accounts you're losing money on, idle waste in dollars, and a repricing shortlist.
Install the read-only agent
Set your API key
Tag your jobs (optional)
See per-job cost + waste
Start with a read-only audit. Once you trust the numbers, NemulAI hands you the break-even rate and a suggested list price per tenant — and flags drift as new accounts slip below cost. You decide what to bill; you stay in control the whole way.
NVML probe, WAL buffer, batched upload. Runs as systemd, Docker, or K8s DaemonSet. Read-only by default.
Ingest, per-job + per-tenant cost attribution, below-cost detection, repricing, chargeback.
Per-tenant cost & margin, chargeback exports, below-cost alerts, weekly owner reports.
You're handing us telemetry, not control. Everything below is true on day one.
Yes. By default it only reads NVML counters and uploads metrics. It never changes your workloads unless you explicitly opt into power-cap autopilot — which has an observation window and automatic rollback.
No. It collects GPU telemetry (utilization, power, memory) and job metadata you tag. Your model weights, datasets, and source never leave your machines.
Yes. The agent is open source — audit it, and point it at your own endpoint. The collector will never be paywalled.
Install is ~2 minutes. You'll see live per-job cost immediately and a full per-tenant billing audit within a 1–2 week pilot.
Drop in a billing export and a usage log. NemulAI attributes every GPU-second, flags the accounts you're losing money on, and suggests where to reprice.
Audit the code, self-host against your own endpoint, or contribute integrations. The agent that collects your data will never be paywalled.
Star on GitHub