claudegpt.fun

Item: claudegpt.fun
Rating: 1.4
Author: FreeAIRouter

Free LLM quota · free models · community reputation · risk review

TypeFree public relay

StatusRisky

Platformopenai-compatible

Reputation28 / 100

Risk level High

Confidence30%

Websitehttps://claudegpt.fun ↗

Discovered fromrouterpark

Overview

claudegpt.fun is an anonymously operated free AI API gateway (relay) advertising free access to Claude/GPT-class large models. Registration is open (no email verification) with an invite-reward system, and payments are currently disabled. Its site subtitle bluntly reads 'this site will just shut down if its users aren't active in the community group,' signaling a casual operator that may close at any time.

Free quota

Mostly free: open registration (registration_enabled=true, no email verification, no Turnstile check), with payment currently disabled (payment_enabled=false). It has an invite-reward scheme (inviter +300, invitee +50 quota points). Exact daily/total quota is undisclosed; /v1/models requires auth (401), so free-quota details are unknown.

Free models

Unknown

Pros

Open self-service registration with no email verification and no Turnstile human check; very low barrier to entry
Claims free access to Claude/GPT-class large models (domain points to Claude Code / GPT use cases)
Multi-format compatible with OpenAI/Anthropic/Google (/v1/models accepts Authorization Bearer, x-api-key, and x-goog-api-key headers)
Has an invite-reward points system that grants quota for referrals

Cons

Operator is fully anonymous: contact_info is empty, with no QQ/TG/email or any contact channel, leaving no recourse if something goes wrong
Terms of service, usage policy, and supported-regions legal documents are all blank (empty content_md); no compliance commitments whatsoever
Offers top-tier closed-source models like Claude/GPT for free, which suggests they may be served via stolen/shared upstream keys; authenticity and compliance are doubtful
No independent community reputation anywhere (V2EX/Linux.do/NodeSeek/Zhihu/GitHub return no discussion of this domain)
/v1/models is visible only after auth; free quota, model list, and rate-limit policy are all opaque
Site subtitle states 'this site will just shut down if its users aren't active in the community group,' an explicit operator admission it may close anytime; stability is extremely poor
Common relay issues: possible sudden exit-scam, rate limiting, conversation data passing in plaintext through a third party; CSP already embeds Stripe/Airwallex, so it could switch to paid at any time

Risk flags

Operator fully anonymous; contact_info empty, no contact channel of any kind
Terms of service / usage policy / supported-regions documents all blank; no compliance or privacy commitments
Offers top-tier Claude/GPT models for free, suggesting stolen or shared upstream keys; high account and data-compliance risk
No independent community reputation anywhere; trustworthiness cannot be verified
Free quota, model list, and rate-limit rules all opaque (/v1/models requires auth)
Subtitle admits 'the site will shut down if community-group users go inactive'; operator explicitly signals it may close anytime, so stability is extremely poor
Conversation data relayed in plaintext through a third party with no privacy guarantee; CSP embeds Stripe/Airwallex, so it may turn paid or exit-scam at any time

Community reputation

No independent community reputation can be found: V2EX, Linux.do, NodeSeek, Zhihu, and GitHub return no usable discussion of claudegpt.fun. The site is operated fully anonymously (no contact info, all legal docs blank); it can only be judged a free AI gateway from its own front-end __APP_CONFIG__. Its subtitle even bluntly reads 'this site will just shut down if its users aren't active in the community group,' reflecting a casual operator that admits it may close at any time. Given the common problems of free top-model proxies (stolen upstream keys, sudden exit), it is assessed strictly as an anonymous free new site.

Recommendation

An anonymous, zero-reputation free relay with blank legal terms — not recommended as a trustworthy free source, and certainly not for production or sensitive use. For curiosity only, use it for small-scale trials, do not top up, do not submit sensitive code or private data, and be ready for it to shut down at any time; its free Claude/GPT likely come from an illegitimate source, so use itself may violate upstream terms of service.

Latest probe

ReachableNo (HTTP -)

Latency809 ms

Platformunknown

Models endpointauth required

Mentions (3)

站点前端配置 window.__APP_CONFIG__ · neutral

site_name=claudegpt.fun；site_subtitle=「本站用户在社区群不活跃就关站咯」；registration_enabled=true、email_verify_enabled=false、turnstile_enabled=false；payment_enabled=false；contact_info 为空；affiliate_enabled=true（邀请返利 inviter=300 / invitee=50）；服务条款等 content_md 全为空。 ↗

网关接口 /v1/models 探测 · neutral

返回 {"code":"API_KEY_REQUIRED"}，接受 Authorization Bearer / x-api-key / x-goog-api-key 三种鉴权头，兼容 OpenAI/Anthropic/Google 格式；模型清单需登录后才可见。 ↗

社区检索结果 · negative

V2EX、Linux.do、NodeSeek、知乎、GitHub 均未检索到 claudegpt.fun 的独立讨论，运营完全匿名、无联系方式、法律条款空白。

Sources

Community discussion · reputation check

Want to verify claudegpt.fun's real reputation before deciding? Read first-hand community discussion:

External discussion quality varies; judge together with the risk review on this page.

FAQ

Is claudegpt.fun safe and trustworthy?

Overall reputation 28/100, risk level High. No independent community reputation can be found: V2EX, Linux.do, NodeSeek, Zhihu, and GitHub return no usable discussion of claudegpt.fun. The site is operated fully anonymously (no contact info, all legal docs blank); it can only be judged a free AI gateway from its own front-end __APP_CONFIG__. Its subtitle even bluntly reads 'this site will just shut down if its users aren't active in the community group,' reflecting a casual operator that admits it may close at any time. Given the common problems of free top-model proxies (stolen upstream keys,…

Will claudegpt.fun shut down? Is the service stable?

Known risks: Operator fully anonymous; contact_info empty, no contact channel of any kind; Terms of service / usage policy / supported-regions documents all blank; no compliance or privacy commitments; Offers top-tier Claude/GPT models for free, suggesting stolen or shared upstream keys; high account and data-compliance risk; No independent community reputation anywhere; trustworthiness cannot be verified; Free quota, model list, and rate-limit rules all opaque (/v1/models requires auth); Subtitle admits 'the site will shut down if community-group users go inactive'; operator explicitly signa…

Which free models does claudegpt.fun offer? How much free quota?

Free quota: Mostly free: open registration (registration_enabled=true, no email verification, no Turnstile check), with payment currently disabled (payment_enabled=false). It has an invite-reward scheme (inviter +300, invitee +50 quota points). Exact daily/total quota is undisclosed; /v1/models requires auth (401), so free-quota details are unknown.. Free models: unknown.

Is claudegpt.fun worth using? How should I use it?