Terms of Service

Last updated: August 11, 2026

These Terms of Service ("Terms") govern your access to and use of FlexInference (the "Service"), including our API, dashboard, and SDKs. By creating an account, accessing our API, or otherwise using the Service, you agree to these Terms. If you are creating or using an account on behalf of an organization, you represent that you have the authority to bind that organization, and "you" refers to both you and the organization.

If you do not agree to these Terms, do not use the Service.

Acceptance of Terms

By accessing or using FlexInference, you agree to be bound by these Terms and by our Privacy Policy, which is incorporated into these Terms by reference. We may update these Terms from time to time, as described in "Changes to the Service or these Terms" below. Continuing to use the Service after an update takes effect means you accept the revised Terms.

Description of Service

FlexInference is a deadline-aware inference router for large language models. You send requests to our API, and we route each one to the provider and service tier that fits the deadline you set.

You can send a request through the "default", "priority", or "auto" tier directly, or through the "flex" tier. Flex attempts to fulfill your request through a lower-cost service option first, within a deadline you set on the request (the "start_within" field). If that flex attempt fails, times out, or will not complete within your deadline, we automatically fall back to the default tier so the request still completes rather than returning an error. Supported providers today are OpenAI, Gemini, and Anthropic, and cloud routes through Amazon Bedrock and Google Vertex AI.

FlexInference does not train or operate its own models. We are a routing and proxying layer that sits between you and the underlying providers.

Accounts and Eligibility

Account creation and login are handled through WorkOS AuthKit. We use it to collect the account information needed to operate the Service, such as your email address, name, and organization membership, and to maintain your session. FlexInference's own code never directly handles your password.

You agree to provide accurate account information and to keep it current. You are responsible for all activity under your account and for keeping your login credentials secure.

You must be at least 18 years old, or the age of majority in your jurisdiction if that is older, to use the Service.

If you create an account on behalf of an organization, you represent that you have authority to bind that organization to these Terms. The organization is responsible for the account and for all activity under it, including activity by other members added to it.

You can run more than one workspace and invite members to a team workspace. Each member's access is limited to the permissions granted to their role. Members with the admin role can manage billing, and only the workspace owner can delete the workspace. If you move a Personal workspace into a team workspace you own, that port transfers your own keys, credentials, wallet balance, usage history, and settings between your own workspaces and leaves the Personal workspace empty.

Bring-Your-Own-Key (BYOK)

FlexInference works on a bring-your-own-key model by default. You provide your own API key for the provider you want us to route to: OpenAI, Gemini, or Anthropic. A cloud route takes that route's own credentials instead, like a Bedrock key and region or a Workers AI token and account id. We use those credentials only to make the requests you send us.

Your agreement with that provider, including your usage limits, billing, and standing with them, is between you and the provider. FlexInference is not a party to that agreement and does not modify it in any way. You remain solely responsible for complying with the provider's own acceptable use policies and other terms for any request we route on your behalf using your key.

We encrypt your provider keys at rest and decrypt them only at the point of making your proxied request. We do not store or log the prompts or completions that pass through the Service. That content is passed through to the provider using your key and is subject to the provider's own policies once it reaches them. See our Privacy Policy for more detail on how we handle keys and content.

Managed Keys

Your organization can switch a provider from BYOK to Managed Keys. On Managed Keys, FlexInference serves that provider on its own pooled credentials, so you do not connect a key of your own for it. You remain responsible for using the Service within the Acceptable Use terms below and the underlying provider's policies.

Managed Keys is billed from a prepaid wallet held in your organization's balance. You top the wallet up through our payment processor, and each top-up adds a platform fee of 10% of the amount you add, with a minimum top-up of $10. We serve managed requests while your balance is positive and settle each request at its real cost. When the balance runs out, managed requests stop until you top up again.

The spendable base of your wallet balance is refundable. On request we return the unused base to the card you paid with. The platform fee is not refundable. We store managed request content only if your organization turns trace storage on, and the optional moderation and PII masking controls process content only when you enable them.

Acceptable Use

You agree not to use the Service to:

• Generate, store, or transmit content that is illegal or that facilitates illegal activity;
• Abuse, disrupt, or attack the Service, including denial-of-service attempts or interference with other customers' use of the Service;
• Attempt to bypass rate limits, authentication, or other security controls on the Service;
• Reverse engineer, decompile, or otherwise attempt to derive the source code or underlying implementation of the Service, except where applicable law prohibits this restriction;
• Use the Service in a way that violates the acceptable use policies of the underlying provider whose key you are using.

We may suspend or terminate access for accounts we reasonably believe are violating this section.

Fees and Billing

The Service offers Bring Your Own Key serving at no charge, with no request allowance and no payment card required. Managed Keys draws model cost from a prepaid wallet you top up, plus one flat top-up fee stated on our site at the time you add funds.

Your dashboard shows your billed cost, an estimate of what the same usage would have cost without FlexInference, and the resulting savings. We may update our pricing and billing approach from time to time; material changes will be reflected on our site and, where required, communicated to you in advance.

Payment, Authorization, and Delinquency

Bring Your Own Key serving is free. Managed Keys is prepaid: you add funds to a wallet, and FlexInference passes model cost through at the provider's rate with no markup. The only FlexInference charge is the flat top-up fee applied when funds are added.

By adding a payment card in the dashboard Billing page, you authorize FlexInference and its payment processor to charge that card for wallet top-ups you initiate. If you enable auto-refill, the same authorization covers automatic top-ups on the settings you choose. You can remove the card or disable auto-refill at any time.

If a top-up payment fails, the balance is not loaded and requests that need it return a payment-required error (HTTP 402) until a payment clears. Serving on your own provider keys is unaffected, because it never draws on the balance. Updating your card and loading again restores managed serving.

All fees are stated and charged in US dollars and are exclusive of taxes. You are responsible for any sales, use, value-added, or similar taxes arising from your use of the Service, other than taxes based on FlexInference's income.

If you believe a charge is incorrect, contact us at [email protected] so we can investigate and, where appropriate, correct it. You agree to raise billing questions with us before initiating a chargeback or payment dispute. Fraudulent or unwarranted chargebacks may result in suspension or termination of your account, and any reversed amounts remain payable together with any related fees we incur.

API Keys and Security

FlexInference issues you an API key to authenticate with our Service, and you separately provide us with your provider keys under BYOK. You are responsible for keeping both your FlexInference API key and your provider keys confidential, and for not sharing them with anyone who should not have access to your account or your provider account.

Your FlexInference API key is shown to you once, at the time you create it. We do not retain or display the plaintext key afterward, so if you lose it, you will need to create a new one.

If you suspect that any of your keys have been compromised or used without authorization, notify us promptly at [email protected] so we can help you secure your account.

Service Availability

We aim to provide reliable, low-latency access to the Service, but we do not guarantee that it will be uninterrupted or error-free. The flex tier in particular depends on the availability and performance of the underlying providers at the time of your request; if a provider is degraded or unavailable, that affects our ability to serve your request through flex within your deadline.

Where a provider returns an error, we pass that error through to you unchanged rather than masking or silently retrying it in a way that hides it from you. We do not guarantee any particular level of uptime, and the Service may be modified, suspended, or become temporarily unavailable, including for maintenance.

Disclaimer of Warranties

THE SERVICE IS PROVIDED "AS IS" AND "AS AVAILABLE," WITHOUT WARRANTIES OF ANY KIND, WHETHER EXPRESS, IMPLIED, OR STATUTORY. TO THE FULLEST EXTENT PERMITTED BY LAW, WE DISCLAIM ALL IMPLIED WARRANTIES, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, AND NON-INFRINGEMENT. WE DO NOT WARRANT THAT THE SERVICE WILL BE UNINTERRUPTED, SECURE, OR ERROR-FREE, OR THAT ANY OUTPUT FROM AN UNDERLYING PROVIDER WILL BE ACCURATE, COMPLETE, OR RELIABLE.

Limitation of Liability

TO THE FULLEST EXTENT PERMITTED BY LAW, FLEXINFERENCE WILL NOT BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, OR FOR ANY LOSS OF PROFITS, REVENUE, DATA, OR GOODWILL, ARISING OUT OF OR RELATED TO YOUR USE OF THE SERVICE, EVEN IF WE HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.

TO THE FULLEST EXTENT PERMITTED BY LAW, OUR TOTAL LIABILITY FOR ANY CLAIM ARISING OUT OF OR RELATED TO THESE TERMS OR THE SERVICE WILL NOT EXCEED THE AMOUNT YOU PAID TO FLEXINFERENCE FOR THE SERVICE IN THE THREE MONTHS BEFORE THE EVENT GIVING RISE TO THE CLAIM.

Some jurisdictions do not allow the exclusion or limitation of certain damages, so some of the limitations above may not apply to you.

Indemnification

You agree to defend, indemnify, and hold FlexInference harmless from any claims, damages, liabilities, and expenses (including reasonable legal fees) arising out of or related to: your use of the Service; your violation of these Terms; your violation of any provider's terms or policies in connection with your use of the Service; or content you generate, submit, or process through the Service.

Termination

You may stop using the Service and close your account at any time. We may suspend or terminate your access if you violate these Terms, if required by law, or if we discontinue the Service, with notice where practical.

On termination, your FlexInference API keys and stored provider keys are revoked and can no longer be used to authenticate against the Service. Any data associated with your account continues to be handled as described in our Privacy Policy.

Provisions of these Terms that by their nature should survive termination, including Fees and Billing (for amounts owed), Payment, Authorization, and Delinquency (for amounts owed and payment authorizations), Disclaimer of Warranties, Limitation of Liability, Indemnification, and Governing Law, survive termination.

Feedback and Submissions

The Service includes ways to send us feedback: bug reports, feature requests, questions, and the contact form. You keep ownership of what you submit. You grant us a non-exclusive, worldwide, royalty-free license to use your submissions to operate, support, and improve the Service. We have no obligation to act on a submission or to compensate you for it.

Do not submit anything illegal, infringing, or malicious. Do not include API keys, passwords, or other secrets in a report or attachment. We may remove submissions or attachments at our discretion. Abuse of these forms, including automated or bulk submission, is a violation of the Acceptable Use section.

Changes to the Service or these Terms

We may change, improve, or discontinue features of the Service at any time. We may also update these Terms from time to time. If we make material changes, we will update the "last updated" date on this page and, where appropriate, provide additional notice. Your continued use of the Service after a change takes effect means you accept the updated Terms.

These Terms were last updated on August 11, 2026.

Governing Law

These Terms are governed by the laws of the State of Delaware, United States, without regard to its conflict of law principles.

Contact Us

If you have questions about these Terms, contact us at [email protected].