No monthly plans. No minimums. Top up credits, call any model, pay the exact per-token rate. Unused balance stays yours.
Three steps. No subscriptions, no auto-renew, no surprise invoices.
Add any amount from $5 up. Stripe charges once. The balance lands on your account immediately.
One sk-maas-… key works across Claude, Gemini, OpenAI, MiniMax. We pass the request through and meter the cost.
Every request shows up in Activity with tokens, latency, and dollars. Run out? Requests 429 — never overcharged.
Pass-through provider prices with a thin managed-platform margin.
| Model | Provider | Input / MTok | Output / MTok |
|---|---|---|---|
Veo 3.1 veo-3.1 | Google AI Studio | — | $0.400 per second |
Gemini 3.1 Flash-Lite gemini-3.1-flash-lite | Google AI Studio | $0.25 | $1.50 |
GPT-5.4 mini gpt-5.4-mini | OpenAI | $0.75 | $4.50 |
Gemini 3.5 Flash gemini-3.5-flash | Google AI Studio | $1.50 | $9.00 |
Nano Banana Pro nano-banana-pro | Google AI Studio | — | $0.134 per image |
Claude Sonnet 4.6 (AWS Bedrock) bedrock-claude-sonnet-4-6 | Amazon Bedrock |
Same product, different rate limits + model access — based on whether you have credits.
No credit card. Just sign up.
Top up any amount. Spend what you spend.
For teams that need SSO, on-prem deployment, custom provider adapters, or reserved upstream capacity — talk to us. Custom contract, not a SKU.
No. Every account is pay-as-you-go. You top up credits when you want to and burn through them at the per-token rates shown above. Nothing recurs.
Text is per-token in + per-token out. Image is per-image. Video is per-second. We use the upstream provider's published rates with a small managed-platform margin baked in — every request shows its exact cost in Activity.
Requests fail with HTTP 429 until you top up. Free-tier keys keep working on open models. You can never be charged more than your top-up balance — there is no postpaid mode.
Not on the managed plan. Miavo runs the upstream credentials so a single sk-maas-… key gets you Claude + Gemini + OpenAI + MiniMax. For full upstream control, the self-host build is the path.
Drop-in. The /v1/chat/completions endpoint speaks the OpenAI schema across all providers — streaming, tool calls, image inputs translated automatically.
| $3.00 |
| $15.00 |
GPT-5.5 gpt-5.5 | OpenAI | $5.00 | $30.00 |
Claude Opus 4.8 (AWS Bedrock) bedrock-claude-opus-4-8 | Amazon Bedrock | $5.00 | $25.00 |
Image and video models are priced per-image or per-second; full breakdown lives on the models page.
Unused balance stays on your account indefinitely. Email us if you ever want it refunded — we don't lock balances in.