← Back to Chat

Terms of Service

Last updated: July 2, 2026

1. What This Is

Unlimited AI ("the Service") is a free, hobby project that provides proxy access to a wide range of third-party AI model APIs. It is hosted on Cloudflare Pages (unlimitedai.pages.dev) and backed by an Abacus.ai RouteLLM subscription that the operator pays for out of pocket. This is not a commercial offering — there is no company, no employees, no investors, and no revenue stream behind it.

2. Why This Service Exists

AI was never meant to be paid — at least not at the prices charged per API call. All of the models accessed through this Service exist as open research, affordable subscriptions, or competitive products whose individual pricing is reasonable. But no one offers them all in one place without charging per-token. This Service fills that gap: one subscription ($10/mo for Abacus.ai RouteLLM), shared freely, with no markup, no usage tiers, and no monetization.

The operator believes access to AI should not be gated by credit card limits. If you find this Service useful, pay it forward — help someone else learn, build, or create with AI. The operator explicitly refuses donations to keep the Service free of financial dependency or obligation.

3. Why Cloudflare Pages (.pages.dev)

The Service runs on Cloudflare Pages for several reasons:

  • Free hosting — No server costs means no financial pressure to monetize
  • Global CDN — Cloudflare serves static assets from 330+ cities worldwide
  • DDoS protection — Built-in protection against attacks
  • HTTPS everywhere — All traffic is encrypted via Cloudflare's TLS
  • Edge functions — API proxying runs at Cloudflare's edge, not a central server
  • No data persistence — Cloudflare Pages Functions are stateless and ephemeral

4. About Malware, Viruses, and Security

This Service poses no malware or virus risk:

  • No executables — The frontend is pure HTML, CSS, and JavaScript. No .exe, .msi, .dmg, .apk, or any executable binaries are served.
  • No third-party scripts — Zero external JavaScript, no ads, no trackers, no analytics, no social media widgets, no embedded frames.
  • Serverless only — Backend logic runs as Cloudflare Pages Functions (JavaScript at the edge). No server processes, no database, no file system.
  • HTTPS delivery — All assets served over TLS from Cloudflare's CDN. Integrity is guaranteed by the TLS certificate chain.
  • Inspectable — Every line of frontend code is visible in your browser's developer tools. The functions source code is minified but functional.
  • Code downloads — When you download AI-generated code via the "Download Code" button, it saves as a plain text file with the appropriate extension (.py, .js, .html, etc.). No executable code is packaged or served.
  • Generated images — Downloadable as standard PNG files. The "Save Image" button uses the browser's native Blob API. No DRM, no watermarking, no hidden data.

5. Data Ownership

Your data is your own. We do not use your data for training LLMs. This is also Abacus.ai's official policy — they do not train on customer data. The Service infrastructure does not log, store, persist, or inspect your prompts or responses beyond forwarding them to the upstream API. There is no database, no analytics, no data broker relationship.

6. Usage Limits

The Service relies on a single Abacus.ai ChatLLM subscription ($10/mo, Basic tier), which provides 20,000 credits per month. Here is how usage works:

  • Without attachments: You can send thousands of text-only messages per month without hitting usage limits on SOTA models like GPT-5.5, Claude Sonnet 4.5, and Gemini 3.1 Pro.
  • With large attachments: Large file uploads consume more credits and will trigger rate limits faster. If you hit a rate limit, the system automatically switches you to a different LLM so you can keep working.
  • Total capacity: Abacus gives roughly 10x more tokens than other LLM services at the same price tier. You get up to ~15M input tokens on expensive models or ~70M tokens on mid-range models per month.
  • Truly unlimited models: RouteLLM, GPT-5 Mini, Gemini 3 Flash, Grok Code Fast, GLM 4.7, Kimi K2, and Llama4 are available without any usage limits — they keep working even after monthly credits are exhausted.
  • The Service is offline when funding runs out. There is no overage, no surprise bill, no credit card on file. When the $10/mo is spent, the Service goes dark until the next billing cycle.

7. Acceptance of Terms

By accessing or using the Service, you agree to be bound by these Terms. If you do not agree, do not use the Service. The operator reserves the right to modify these terms at any time without notice.

8. No Warranties

The Service is provided without any warranty, express or implied. The operator does not guarantee: (a) uninterrupted or error-free operation, (b) the accuracy, reliability, or completeness of AI-generated content, (c) that the Service will meet your requirements, (d) that the Service will remain available indefinitely.

9. Limitation of Liability

Under no circumstances shall the operator be liable for any damages arising from the use or inability to use the Service, including but not limited to direct, indirect, incidental, consequential, or punitive damages. The AI models accessed through this Service are provided by third parties (OpenAI, Anthropic, Google, xAI, DeepSeek, Alibaba, Meta, Zhipu AI, MiniMax, and others). The operator has no control over the output of these models and assumes no responsibility for generated content.

10. Prohibited Uses

You agree not to use the Service for: (a) any illegal activity, (b) generating harmful, abusive, or violent content, (c) attempting to circumvent rate limits or security measures, (d) reverse-engineering the Service infrastructure, (e) automated scraping or data mining, (f) any activity that could harm minors.

11. Rate Limiting

The Service employs rate limiting (120 requests per minute per IP) to prevent abuse. The operator may modify rate limits at any time. Attempting to bypass rate limits may result in permanent banning.

12. Third-Party Services

This Service proxies requests to Abacus.ai's RouteLLM API, which in turn routes to specific third-party AI model providers including but not limited to OpenAI, Anthropic, Google, xAI, DeepSeek, Alibaba Cloud, Meta, Zhipu AI, and MiniMax. Your use of these services through this proxy is subject to their respective terms of service and privacy policies.

13. Service Availability

The Service is funded by a single $10/mo Abacus.ai subscription paid by the operator. When the subscription renews each month, the Service comes back online. If the operator stops funding it, the Service goes offline permanently. There is no SLA, no uptime guarantee, no support team. Use it while it lasts — do not build dependencies on its continued existence.

14. No Affiliation

This Service is not affiliated with, endorsed by, or connected to Abacus.ai, OpenAI, Anthropic, Google, xAI, DeepSeek, or any other AI model provider. All trademarks are property of their respective owners.

15. No Donations

The operator explicitly refuses donations, payments, or any form of financial support from users. There is no Patreon, Ko-fi, Buy Me a Coffee, cryptocurrency address, or payment processor associated with this Service. This is a deliberate choice to keep the Service free of any financial obligation or expectation.

16. Indemnification

You agree to indemnify and hold harmless the operator from any claims, damages, or expenses arising from your use of the Service or violation of these Terms.

17. Governing Law

These Terms shall be governed by the laws of the United States. Any disputes shall be resolved in the courts of the United States.

18. Contact

This is an anonymous hobby project. There is no formal support channel. To send feedback, use the Feedback button in the sidebar of the app.