Getting started

From sign-up to first token in three steps.

Your requests run on private, US-hosted GPUs. Nothing you send is used to train shared models, and your prompts and outputs never leave your tenant.

1 · Create your workspace

Start a free trial — no card required. You'll land in your tenant portal with a one-time primary access key (it looks like sov-…). Copy it somewhere safe; it's shown once. You can mint additional scoped keys (delegate / read-only) from the portal's API keys tab.

2 · See what models are available

Every request authenticates with your key as a Bearer token. List the models your workspace can reach:

# replace with your API host and key
curl https://api.nexusaisolutions.io/llm/catalog \
  -H "Authorization: Bearer sov-YOUR_KEY"

3 · Make your first request

Chat completions stream back token-by-token over Server-Sent Events. Pick a backend and model from the catalog:

curl -N https://api.nexusaisolutions.io/llm/chat \
  -H "Authorization: Bearer sov-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "backend": "local-direct",
    "model": "qwen2.5-7b-instruct",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Summarize this contract clause in plain English."}
    ],
    "temperature": 0.7,
    "max_tokens": 512
  }'

The response is an SSE stream of OpenAI-shaped chunks (choices[0].delta.content). Read it incrementally, or buffer to the final message. The backend/model values come from /llm/catalog; local-direct is the built-in private model on your tenant.

Billing & usage

Usage is metered in GPU-minutes — the time a GPU spends on your work. A bigger or heavier model simply consumes more GPU-minutes per request, so one meter covers every model. Each plan includes a monthly bucket:

Plan	Price	GPU-min / mo	Seats	Concurrency
Trial	$0 · 7 days	500	1	1
Solo	$39	3,000	1	2
Studio	$149	15,000	5	4
Team	$599	60,000	20	12
Enterprise	from $2,500	dedicated	custom	dedicated

Past your monthly bucket you can auto-upgrade to the next tier (the default) or pay overage at $0.02 / GPU-minute — your choice, set on your account. Check your current usage any time:

curl https://api.nexusaisolutions.io/billing/YOUR_TID/summary \
  -H "Authorization: Bearer sov-YOUR_KEY"

Keeping your data private

• Requests run on GPUs reserved to your tenant's isolation boundary; no prompt text or output is written to the billing ledger.
• We never train shared models on your data. Fine-tuning happens only on your instruction and stays in your tenant.
• Compute is US-hosted. Regulated firms can run dedicated or on-prem — talk to us about a BAA or GLBA agreement first.

Support

Solo: email. Studio: email + shared Slack. Team: priority support with an 8-hour SLA. Enterprise: named contact and a custom SLA. Questions any time: contact us.

API host shown as api.nexusaisolutions.io for illustration — use the host provided with your workspace.