Terms of Service

Version 4 · Last updated 22 October 2026

These Terms of Service (“Terms”) are an agreement between you and Five S LLC, the operator of MultiO (“MultiO”, “we”, “us”) for the use of the MultiO service at multio.ai (the “Service”). By creating an account or using the Service, you agree to these Terms and to our Privacy Policy.

Operator legal name: Five S LLC. [Operator address — to be inserted before public launch]

1. What MultiO is

MultiO is a subscription service that gives you access, through one interface and one account, to AI models operated by third-party providers — currently the model families offered by OpenAI, Anthropic, Google, and OpenRouter, which routes requests to hosted model providers including Meta's Llama models and xAI's Grok models. MultiO adds its own product layer on top: automatic routing between models, cross-model memory, side-by-side comparison, and metering in MultiO tokens at published rates. Model attributions and licences are listed on our Attributions page.

MultiO is built on the providers’ commercial application programming interfaces (APIs). We do not operate the models, and the models are not ours. Section 10 explains what that means for availability.

2. Eligibility and your account

• You must be at least 18 years old to use MultiO.
• You are responsible for the accuracy of the information you give us, for keeping your password confidential, and for all activity under your account.
• One account is for one person. Do not share your account credentials, and do not resell or transfer your access.

3. Plans, tokens, and how metering works

This section is the heart of the product, and it is written to match the meter exactly.

• Free trial. $0, including a grant of 3,000,000 MultiO tokens, on small and mid-tier models.
• Pro. $20 per month, including a grant of 80,000,000 tokens each month, on all model tiers.
• Family seats. +$12 per month per seat. Each seat is its own account, linked to the payer, with its own 80,000,000-token monthly grant. The payer is responsible for the seats they add.

Metering

• Every message is charged in MultiO tokens at the model’s published token rate. Token rates follow the model providers’ published list prices, converted at a fixed pivot: 1,000,000 MultiO tokens = $0.15 of standard-tier usage. Every model’s rate — MultiO tokens per 1,000 real input tokens and per 1,000 real output tokens — is listed on the Token rate card in the app (Usage & billing) and on the pricing page. The exact charge of every message is shown on the message itself (“the receipt”), together with the real tokens the model used and your remaining balance. There are no hidden deductions.
• No silent revaluation. We will not change token rates, the pivot, or grant sizes for your subscription without the advance notice promised in Section 5 (at least 7 days, and no more than 30 days, before the change takes effect), and a change never applies retroactively to tokens already in your balance.
• Charges are drawn in this order: (1) your monthly token grant, (2) top-up tokens, (3) referral token credits.
• Your monthly token grant resets at the start of each monthly period. Unused grant expires at the reset and does not roll over. Purchased top-up tokens and referral token credits do not expire while your account is active.
• Hard stop. When your token balance reaches zero, the Service stops sending messages until the next reset or a top-up. We never charge overages after the fact and there are no surprise bills: the most a plan can consume is the token grant included in it plus any top-up tokens you chose to buy.
• Web search. Turning web search on adds a flat charge of 100,000 tokens per search on top of the message’s token charge; it is shown as its own line on the receipt. On the free trial, web search is limited to 10 searches per day and cannot be combined with flagship models.
• Cache. An identical repeat question may be answered from cache at a charge of 0 tokens, and the receipt says so.
• Token grants, top-ups and credits are a usage meter. They are not money, carry no cash value, are non-transferable, and cannot be redeemed for cash.

4. Top-ups

Top-ups are one-time purchases of additional tokens: US$5 = 30,000,000 tokens, US$10 = 60,000,000 tokens, US$25 = 150,000,000 tokens at launch. They are spent only after your monthly token grant, and they do not expire while your account is active. Section 6 covers refunds.

5. Billing, renewal and cancellation

• Paid plans are billed in advance through our payment processor, Stripe. Card details are collected and stored by Stripe; they never touch our servers.
• Your subscription renews automatically at the end of each period until you cancel.
• You may cancel at any time. Cancellation stops the next renewal; your plan stays active until the end of the period you have paid for.
• We may change plan prices, token grants, token rates, or the models included in a plan. We will give notice of such a change at least 7 days, and no more than 30 days, before it takes effect for your subscription, together with how to cancel if you do not accept it. A change applies from your next renewal after that notice, never retroactively.

6. Refunds

• Duplicate or erroneous charges are always refunded.
• Except where required by law or the point above, subscription fees for a period that has already begun are not refundable, and the unused monthly token grant does not convert to cash or carry into a later period.
• A top-up that is entirely unused may be refunded on request within 30 days of purchase. Once any part of a top-up has been spent, it is not refundable.
• If we permanently discontinue the Service, we will refund prepaid subscription fees pro rata for the unused remainder of the current period.

If you are a consumer in the EEA or UK: you may withdraw from a new subscription within 14 days of purchase for a full refund, unless you have used the service during that period with your express consent to immediate performance and acknowledgment that the withdrawal right is then lost, as the law provides. Nothing in this section limits any mandatory consumer right.

7. Acceptable use

Because MultiO sends your requests to third-party model providers under their commercial terms, the providers’ usage rules flow down to you. By using the Service you agree:

• to comply with the law and with the providers’ usage policies for the models you use, including the OpenAI usage policies, Anthropic's Usage Policy, Google's Prohibited Use Policy, the Llama Acceptable Use Policy, and xAI's Acceptable Use Policy, each as published by the provider;
• not to use the Service for anything illegal, including generating or storing child sexual abuse material (zero tolerance — such material is reported to the appropriate authorities);
• not to use the Service to create malware, to attack or probe systems you do not own, to extract anyone’s credentials or API keys, or to deceive or defraud;
• not to circumvent metering, rate limits, region restrictions or other Service controls, and not to automate abusive volumes of signups or requests;
• not to resell the Service, share your account, or use the Service to provide API access to third parties;
• not to use outputs from the Service to develop a model that competes with the model that produced them (a condition of the providers’ own terms).

We are answerable to the providers for the activity run under our accounts, so we enforce this section: we may rate-limit, suspend, or terminate accounts that violate it, and we may refuse or re-route requests where a provider’s policy requires it.

8. AI outputs

Outputs are generated by third-party models and may be wrong, incomplete, or outdated, even when they sound confident. Outputs are not professional advice of any kind — not medical, legal, financial, or otherwise. You are responsible for how you use outputs.

As between you and us, you own your inputs and, to the extent it is assignable, the outputs you generate, subject to the providers’ terms. Outputs may not be unique: other users may receive similar outputs.

9. Your content and share links

You keep your rights in the content you submit. You grant us the licence needed to transmit, process and store it in order to operate the Service — including sending it to the model providers described in the Privacy Policy.

Share links make a conversation or comparison visible to anyone who has the link until you revoke it. You are responsible for what you share. Do not use share links to republish other people’s copyrighted text or anyone’s private information (see our DMCA page).

9A. MultiO Code: repositories and pull requests

MultiO Code is an optional feature, available on the plans that include it, that connects MultiO to repositories you choose on GitHub. This section applies when you use it; the rest of these Terms, including the metering rules in Section 3 and the acceptable-use rules in Section 7, apply as well.

• How the connection works. You connect a repository by installing the MultiO Code GitHub App on your GitHub account or organisation and choosing which repositories it may access. MultiO can only ever see the repositories covered by that installation. You may only connect repositories you own or are authorised to grant access to. You can narrow or remove the access at any time on GitHub, or disconnect in MultiO's Settings; either action ends our access.
• What the agent does. In a Code session, an AI model reads files from a snapshot of the repository and branch you select, and proposes changes. Each proposal is shown to you as a per-file diff. Nothing is sent to GitHub until you approve it. When you apply approved files, we create one new branch (named multio/…), commit exactly the content you approved to that branch, and open one pull request. Pull requests are authored by the MultiO Code app (multio-code[bot]) and name you as the requester.
• We never merge. MultiO does not merge pull requests, does not commit to your default branch, and does not change any branch other than the one it creates for a pull request you approved. Reviewing a pull request, resolving conflicts, running your own checks, and deciding whether to merge it all happen on GitHub, by you. A pull request is a request: the repository's owners and rules decide what lands.
• Your responsibilities. You are responsible for reviewing every diff before you approve it and every pull request before you merge it; for the permissions you grant, including access other people have through your GitHub account or organisation; for the code and other content in your repositories, including respecting the licences and rights of anyone whose work the repository contains; and for anything that happens in your repository after a change is merged. Do not use MultiO Code to propose changes you would not have the right to make directly.
• Phase 1 does not run your code. In this phase, MultiO Code reads and proposes text only. It does not execute, build, install, or test any code — yours or its own proposals. Proposed changes are therefore untested by us, may not compile or work, and need your own review and testing like any other draft code. Section 8 (AI outputs) applies to proposals in full.
• Metering. Code sessions are charged in MultiO tokens at the model's published rate, like any other message: each model call in a session is charged at the rate on the Token rate card, the charge is recorded in the same metering ledger, and the tokens a session has used are shown in the session itself. Before a session step runs, we check the worst-case cost against your balance and refuse the step rather than overdraw it. Reading, searching and diffing files on our side is not charged — only model calls are. A session has a hard cap of 4,000,000 tokens; when the cap is reached the session stops, and any proposals already made remain available for your review. The hard stop in Section 3 applies: at a zero balance, sessions stop too.
• Abuse. Using MultiO Code to attack, probe, or degrade repositories or systems; to propose malware or credential-stealing code; to circumvent the session caps, rate limits or the file restrictions described in the Privacy Policy; or otherwise in breach of Section 7 or GitHub's terms, may lead us to suspend MultiO Code, or your account, under Sections 7 and 13.
• GitHub's role. MultiO Code works through GitHub's API and depends on GitHub's availability. Your use of GitHub remains governed by GitHub's own terms, which we do not control.

10. Availability, regions and changes

• The Service depends on third-party model APIs. Providers may add, remove, reprice, restrict or discontinue models at any time, and a model’s availability may differ by country. Where a provider restricts its models by region, we enforce that restriction — normally by re-routing your request to an available model family, and by refusing the request if no alternative exists.
• We may change, suspend or discontinue features. We will give reasonable notice of material changes where we can.
• We work to keep the Service available, but we do not guarantee uninterrupted or error-free operation.

11. Copyright

We respond to notices of alleged copyright infringement under the Digital Millennium Copyright Act as described on our DMCA page. It is our policy to terminate, in appropriate circumstances, the accounts of users who are repeat infringers.

12. Disclaimers and limitation of liability

[Version 1 boilerplate — to be reviewed before public launch]

• The Service is provided “as is” and “as available”, without warranties of any kind, express or implied, to the maximum extent permitted by law.
• To the maximum extent permitted by law, our total liability arising out of or relating to the Service is limited to the greater of (a) the amounts you paid us in the twelve months before the claim, or (b) US$100.
• To the maximum extent permitted by law, we are not liable for indirect, incidental, special, consequential or punitive damages, or for lost profits or data.
• Nothing in these Terms limits liability that cannot be limited under applicable law.

13. Termination

You may stop using the Service and request account closure at any time (see the Privacy Policy for how deletion works). We may suspend or terminate your access if you breach these Terms or a provider policy that flows down under Section 7. Sections 6, 8, 9, 12 and 14 survive termination.

14. Governing law

[Version 1 placeholder — governing law and venue to be confirmed with the operating entity before public launch]

These Terms are governed by the laws identified in the final published version, and any disputes will be resolved in the courts identified there. If any provision of these Terms is held unenforceable, the remaining provisions continue in full effect.

15. Changes to these Terms

We may update these Terms. Material changes will be posted on this page with a new version number and date, and, where the change affects billing, will apply from the next renewal after notice. Continuing to use the Service after a change takes effect means you accept it.

16. Contact

Questions about these Terms: legal@multio.ai.