A GitHub repository, free-LLM, compiles permanently free LLM APIs, free models, and limited-time free tiers from Chinese model vendors and domestic AI platforms. It targets Chinese developers seeking official, directly reachable API endpoints rather than third-party wrappers, and is MIT-licensed.
A curated directory of LLM relay/aggregator gateways that offer free quotas, helping users find which gateways provide free usage and which models are free on each. It includes model specifications and provides static JSON endpoints for programmatic access.
A curated directory of LLM relay/aggregator gateways that offer free quotas, helping users find which gateways provide free usage and which models are free on each. It includes model specifications and provides static JSON endpoints for programmatic access.
213 is a platform that aggregates over 300 free AI models from more than 50 providers, allowing users to access them all through a single interface. It eliminates the need for complex API keys and setups, with smart routing and automatic failover to ensure reliability. Currently in beta, it aims to simplify AI model usage for developers and enthusiasts.
213 is a platform that aggregates over 300 free AI models from more than 50 providers, allowing users to access them all through a single interface. It eliminates the need for complex API keys and setups, with smart routing and automatic failover to ensure reliability. Currently in beta, it aims to simplify AI model usage for developers and enthusiasts.
SingularityAPI provides a single API key and base URL to access multiple open-weight LLMs including DeepSeek V4 and Kimi K2. It is OpenAI-compatible, supports streaming, and offers free credits to beta users. The service targets developers who want to experiment with various models without managing separate accounts.
SingularityAPI provides a single API key and base URL to access multiple open-weight LLMs including DeepSeek V4 and Kimi K2. It is OpenAI-compatible, supports streaming, and offers free credits to beta users. The service targets developers who want to experiment with various models without managing separate accounts.
A Medium article provides a comprehensive overview and comparison of Zhipu AI's GLM model family, which originated from Tsinghua University. The guide covers the models' open-weight availability under permissive licenses, making them accessible for developers and researchers.
A Medium article provides a comprehensive overview and comparison of Zhipu AI's GLM model family, which originated from Tsinghua University. The guide covers the models' open-weight availability under permissive licenses, making them accessible for developers and researchers.
TensorMux launched a free API endpoint providing access to open models like Llama 3.1, Qwen3, GPT-OSS, and Gemma. Users receive an API key and $5 of free credit (1 million tokens) without requiring a credit card. The service is currently in a shared pool with rate limits and a small model list.
TensorMux launched a free API endpoint providing access to open models like Llama 3.1, Qwen3, GPT-OSS, and Gemma. Users receive an API key and $5 of free credit (1 million tokens) without requiring a credit card. The service is currently in a shared pool with rate limits and a small model list.