Pricing

    Start free. Pay for what you run.

    No seats, no minimums, no sales call.

    Free

    $0
    no card required
    • 1,000 runs / month
    • 20 requests / minute
    • Every model and tool
    • Knowledge and memory included
    • Community support
    Get started free

    Pay as you go

    $0.07
    per run · card payment, no call
    • Unlimited runs
    • 300 requests / minute
    • Spend caps and usage alerts
    • LLM tokens billed separately
    • Email support
    Add a card

    Enterprise

    Volume
    dedicated infrastructure
    • Volume pricing
    • Custom rate limits
    • Custom terms, SLA and DPA
    • Security review support
    • Named technical contact
    Talk to us

    Definition

    What counts as a run.

    One request that triggers three tool calls and a knowledge lookup is one run. Steps are not billed; tasks are.

    1 request1 model call3 tool calls1 knowledge lookup1 retry1 guardrail check
    = one run, $0.07

    LLM costs

    On your key, on your account.

    Bring your own provider key and model tokens bill directly to you at your own negotiated rate. There is no markup, because the spend never reaches us.

    That leaves one GTWY line: the run fee, taken from your prepaid wallet as each run executes.

    See per-model rates →

    Free tier

    Exactly what you get.

    Runs1,000 / month
    Rate limit20 requests / minute
    Card requiredNo
    ModelsAll of them
    Tools and knowledgeIncluded
    Not includedSLA, dedicated infrastructure

    Free-tier limits are proposed, not approved — see the note above.

    Calculator

    What your month looks like.

    Runs plus tokens. No resale maths here — you’re building this for yourself.

    Free tier covers1,000 runs
    Billable runs19,000 runs
    Run fees$1,330.00
    Token costs$720.00
    Estimated total$2,050.00 / mo

    Token cost uses this model’s published rate at a 50/50 input–output split. Your real split will differ.

    Questions

    Before you put a card on it.

    What happens when I hit the free limit?

    Runs stop until the month resets, or until you add a card. Nothing is deleted and no run is silently billed — you get a 402 with a clear reason rather than a surprise charge.

    Are tool calls charged separately?

    No. Tool calls, knowledge lookups, retries and fallbacks all sit inside the run you already paid for.

    Are failed runs charged?

    A run that fails on our side is not billed. A run that completes and returns something you didn’t like is billed — the work happened.

    Can I set a spend cap?

    Yes. Set a monthly ceiling and an alert threshold; runs stop at the cap rather than continuing past it.

    Do you mark up LLM costs?

    No — because we never handle them. Tokens bill to your own provider account on your own key, at whatever rate you have with them. The only thing GTWY charges is the run fee.

    Can I use my own model key?

    Yes. Add your key and inference runs on your account. You will also need a wallet balance for the run fees, which are deducted per run as they execute.

    Start on the free tier.

    1,000 runs a month, no card. Add one when you outgrow it.