Use Policy

1. Purpose StringRoute's value pricing does not mean unrestricted use of every advanced model or permission to bypass risk controls, rates, concurrency, context, file-size, regional, or content-safety limits.

2. Rules for Webchat users Do not share one account among multiple people or use it for teams, resale, question brokering, or bulk generation. Do not use scripts, automation tools, browser clusters, or proxy pools for bulk calls. Do not consume advanced models, file analysis, image understanding, or other high-cost features continuously, at abnormal frequency, or in an abusive manner. Do not upload unmasked sensitive information, third-party private data, trade secrets, or regulated data. Do not occupy shared resources by repeatedly refreshing, switching accounts, or evading limits. We may impose per-day, per-hour, per-minute, or per-request Webchat limits, including message counts, tokens, file size, image uses, context length, model availability, and queue priority.

3. Rules for API users Do not share, sell, publish, or expose an API Key in frontend code. Do not use the API for spam, phishing, fraud, malicious code, automated attacks, credential stuffing, CAPTCHA bypass, fake reviews, or other unlawful content or conduct. Do not evade rate limits or risk controls through abnormal concurrency, proxy pools, multiple accounts, or cycling top-ups. Do not use the API in ways that violate upstream model-provider policies. Do not submit personal or sensitive data without required authorization from the end user. We may set QPS, RPM, TPM, concurrency, model allowlists, per-request context, retry rules, and regional limits based on account reputation, balance, plan, call history, risk level, and upstream limits.

4. Dynamic adjustments We may adjust available models, routing, pricing, multipliers, token billing, free quotas, membership benefits, request rates, concurrency, queues, files, images, context, and batch limits based on system load, upstream cost, model availability, abnormal risk, and policy requirements. If an adjustment affects purchased benefits, we will use reasonable efforts to explain it through the website, dashboard, or email. We may act first when required by an urgent upstream change, security event, unlawful risk, or force majeure.

5. Risk-control measures and appeals If the system detects abnormal conduct, we may require verification of identity, email, payment information, or intended use. We may temporarily rate-limit or queue requests, restrict models, lower priority, suspend an API Key, freeze an account or balance, reject high-risk requests, or block prohibited content. For serious violations, we may terminate service, deny refunds, and retain logs for security, audit, dispute, and legal purposes. To appeal a restriction or suspension, email [email protected] with the account, order, last four characters of the API Key, request time, error details, and business context. Target review period: 10 business days.