api.llm7.io

Item: api.llm7.io
Rating: 3.3
Author: FreeAIRouter

免费大模型额度 · 免费模型 · 社区口碑 · 风险评估

网站类型公益站(中转)

状态可用

口碑分66 / 100

风险等级中

置信度70%

官网https://llm7.io ↗

发现来源for-the-zero/Free-LLM-Collection

网站介绍

LLM7.io 是一个由个人/社区驱动（AGPL-3.0 开源、捐赠支持）的免费 LLM API 聚合网关，提供 OpenAI 兼容接口（api.llm7.io/v1），聚合 OpenAI、DeepSeek、Mistral、Gemini 等多家上游模型，匿名即可调用。

免费额度

免费层由捐赠支持：匿名约 30 RPM（每分钟请求），在 token.llm7.io 免费注册领取 token 后提升到 120 RPM；约 2 req/s、20/min、100/hour（文本端点，按文档口径）。另有付费 Pro 计划。不保证可用性与额度。

免费模型

gpt-4o-minideepseek-r1-0528deepseek-v3-0324gemini-2.5-flash-litemistral-small-3.1-24bqwen2.5-coder-32b（以及约 24+ 其它模型）

优点

提供真正可编程调用的免费 API：OpenAI 兼容 base URL api.llm7.io/v1，匿名即可用（api_key 可填 'unused'）
免费提供 GPT-4o-mini 等较稀缺模型，无需信用卡、无需付费 key
开源（AGPL-3.0）、GitHub 公开仓库，运营相对透明，注册免费拿 token 提速简单
聚合多家上游，模型选择丰富

缺点

第三方网关，非 OpenAI/DeepSeek/Meta 官方，模型 key 来源与合规性需自行评估
官方声明不保证正常运行时间、延迟、模型可用性与输出准确性，可随时下线/限制
由个人运营（support@llm7.io），抗风险与长期可持续性弱于机构
免费层速率较低（匿名 30 RPM），重度/生产使用受限
多家上游模型的免费供给来源不透明，存在'借用上游免费/试用层'的潜在风险

风险点

第三方聚合网关，非任一模型官方，上游免费/key 来源与合规性不透明
官方明确不保证可用性/延迟/模型存续，可随时撤下模型与功能
由个人运营，长期可持续性与跑路风险需自行评估
免费层速率低（匿名 30 RPM / token 后 120 RPM），不适合生产或高并发
数据合规：经第三方转发，勿提交敏感/隐私内容

社区口碑综述

在多份英文免费 LLM API 清单（freellm.net、cheahjs/free-llm-api-resources、mnfst/awesome-free-llm-apis 等）中被收录并整体正面评价，亮点是免费 GPT-4o-mini 访问、注册简单无信用卡。但同一批来源也明确提醒：它是第三方网关而非官方供应方，对'声称提供官方免费 key'的服务要保持警惕；官方条款也强调不保证可用性。社区中文论坛讨论较少。

使用建议

适合个人开发者免费做实验、原型与轻量调用：用 api.llm7.io/v1 OpenAI 兼容接口，匿名即可，在 token.llm7.io 领免费 token 提速。但因属第三方网关、官方不保证可用性、由个人运营，不建议用于生产或敏感数据场景，使用前先小规模试用。

社区提及 (9)

freellm.net (Niche Free LLM Providers) · positive

Free. One of the few places to get OpenAI's own models without a paid key. 30 RPM (120 with token registration). Standout is free GPT-4o-mini access, rare on a free tier. Simple registration, no credit card. ↗

Medium (Eugene Evstafev / chigwel) · positive

Free and anonymous access to state-of-the-art models without API tokens or authentication; base URL https://api.llm7.io/v1 with api_key set to 'unused'; 30 RPM raised to 120 with a free token from token.llm7.io. ↗

freellm.net (Free LLM API providers list) · neutral

Critically, LLM7.io is not affiliated with the official providers of the models (OpenAI/DeepSeek/Meta). It's a third-party gateway, not an official provider — be cautious of any service claiming free official keys. ↗

GitHub chigwell/llm7.io README · neutral

Community-driven open-source project (AGPL-3.0), powered by Azure, Cloudflare, DeepSeek, Ollama/LLaMA, Mistral, Nebius, Nebula Block, OpenAI, Pollinations, Scaleway; not an official service from any single AI provider. ↗

Grokipedia / 第三方词条 · neutral

Operated by an individual (support@llm7.io), based in the UK and complying with UK/EU GDPR; offers Default/Fast/Pro routing modes; free plan ~2 req/s, 20/min, 100/hour for text endpoints. ↗

freellm.net (Niche Free LLM Providers) · positive

Free. One of the few places to get OpenAI's own models without a paid key. 30 RPM (120 with token registration). Standout is free GPT-4o-mini access. Simple registration, no credit card. ↗

freellm.net (Free LLM API providers list) · neutral

Critically, LLM7.io is not affiliated with the official providers of the models. It's a third-party gateway, not an official provider — be cautious of any service claiming free official keys. ↗

GitHub chigwell/llm7.io README · neutral

Community-driven open-source project (AGPL-3.0), powered by Azure, Cloudflare, DeepSeek, Mistral, OpenAI and others; not an official service from any single AI provider. ↗

Medium (Eugene Evstafev / chigwel) · positive

Free and anonymous access to SOTA models; base URL https://api.llm7.io/v1 with api_key set to 'unused'; 30 RPM raised to 120 with a free token from token.llm7.io. ↗

参考来源

In English

Summary: LLM7.io is an individual/community-driven (AGPL-3.0 open-source, donation-supported) free LLM API aggregation gateway offering an OpenAI-compatible endpoint (api.llm7.io/v1), aggregating upstream models from OpenAI, DeepSeek, Mistral, Gemini and others, callable even anonymously.

Free quota: Donation-supported free tier: ~30 RPM anonymously, raised to 120 RPM after registering for a free token at token.llm7.io; roughly 2 req/s, 20/min, 100/hour for text endpoints per docs. A paid Pro plan also exists. Availability and quota are not guaranteed.