Build, run and scale AI products from one control plane. One gateway for every model, billing your auditors can read, and keys your security team will sign off.
Model gateway · Metered credits · Scoped keys · Audit trail
checking gateway status# One request. Routed, metered, settled, audited. $ curl https://vybeplane.ai/gateway/v1/chat/completions \ -H "authorization: Bearer vp_live_d4TNyN…" \ -d '{"model":"vybe-mock/echo-small","messages":[{"role":"user","content":"Hello"}]}' { "id": "req_2bqoHbwBKykgkGfCKuBjzK", "provider": "vybe-mock", "usage": { "total_tokens": 16, "cost_micro_credits": 3 } } # Settled once on the ledger · balance 4,999,997 · audit event recorded
VybePlane replaces the glue code every AI company rewrites — routing, metering, key management, audit — with one platform your whole team can operate.
One OpenAI-compatible endpoint for every model. Automatic provider fallback with per-request attempt reporting — your product never sees a single-provider outage.
Billing you can audit. Every charge is an immutable ledger transaction — balances are derived, never mutated, and duplicate settlements are impossible by design.
Workspace-scoped API keys, hashed at rest and shown exactly once. Scoped permissions, one-call revocation, and cross-tenant access that simply does not resolve.
Every request accounted for. Usage records tie each call to its ledger transaction, and an append-only audit trail covers every workspace action.
Publish real pages and sites to the internet in one call — or one tap in the quickstart above. Every deploy is metered on your ledger and written to your audit trail. Container apps are next.
Managed PostgreSQL provisioned beside your workloads, with authentication, storage and realtime on the same roadmap.
No terminal, no credit card. Create your workspace, ask the AI, and put a real page on the internet — right here on this screen. Every new workspace includes 5 free credits.
This is your private space on VybePlane. Nothing to install.
Free to start · 5 credits included · no card needed
They are shown once, like a hotel safe code. Copy both somewhere safe — a note on your phone is fine.
Your first metered request — routed, priced and settled on your ledger in real time.
Give it a name and press publish. It goes live on the internet in about a second.
Your page is on the internet, served by VybeDeploy.
Open my pageWorkspace created · keys issued · AI call metered · page published
Balance micro-credits · every charge is on your ledger
# 1. Create a workspace
curl -s https://vybeplane.ai/v1/workspaces -H 'content-type: application/json' \
-d '{"name":"My Workspace","ownerEmail":"you@example.com"}'
# 2. Mint an API key (management token from step 1)
curl -s https://vybeplane.ai/v1/workspaces/WS_ID/api-keys \
-H 'authorization: Bearer MGMT_TOKEN' -H 'content-type: application/json' \
-d '{"name":"my key","scopes":["gateway.invoke"]}'
# 3. Call the gateway
curl -s https://vybeplane.ai/gateway/v1/chat/completions \
-H 'authorization: Bearer VP_LIVE_KEY' -H 'content-type: application/json' \
-d '{"model":"vybe-mock/echo-small","messages":[{"role":"user","content":"Hello"}]}'
# 4. Deploy a site
curl -s https://vybeplane.ai/v1/workspaces/WS_ID/sites \
-H 'authorization: Bearer MGMT_TOKEN' -H 'content-type: application/json' \
-d '{"name":"my-site","files":[{"path":"index.html","content":"<h1>Live</h1>"}]}'
Fetched from the gateway as you read this. The developer preview ships deterministic mock models for integration work; frontier providers land next on the roadmap.
| Model | Providers | Prompt / 1k tokens | Completion / 1k tokens |
|---|---|---|---|
| loading from /gateway/v1/models… | |||
No seats, no minimums. Usage settles against your wallet at request time — every debit is a ledger transaction you can inspect.
Free with every new workspace. 1 credit = 1,000,000 micro-credits — enough for extensive integration testing against the preview catalogue.
Priced per 1,000 tokens per model, rounded up per request. Exact rates are in the live catalogue above — what you see is what settles.
Volume pricing, custom routing policies and dedicated capacity as the platform grows. Write to hello@vyberoom.ai.
Checked from your browser against production, right now.