Free AI Router · free LLM API radar | risk review of free AI API sites

// STATIONS 213 sites · page 1/9

Official (free tier)Up

huggingface.co

Source · cheahjs/free-llm-api-resources

OverviewHugging Face is a leading open-model community and inference platform offering an OpenAI-compatible API via Inference Providers; each user gets small monthly free credits to call many open models.

Free quotaMonthly free credits: ~$0.10 for Free users (official notes subject to change), $2.00 for PRO, $2.00/seat for Team/Enterprise; Serverless Inference has free allowance for models under 10GB (some popular larger models also supported).

Free modelsLlama 系列 Qwen 系列 Gemma 系列 Mistral 系列 DeepSeek 系列众多 HuggingFace Hub 开源模型

See huggingface.co's free quota / risk review →huggingface.co

Official (free tier)Up

console.groq.com

Source · cheahjs/free-llm-api-resources

OverviewGroq's official developer console, famed for ultra-fast inference, offers an OpenAI-compatible API with a no-time-limit, no-credit-card free tier serving ~a dozen open LLMs.

Free quotaFree tier has no time limit and no credit card: e.g. Llama 3.1 8B ~14,400 req/day, 6,000 tokens/min; Llama 3.3 70B ~1,000 req/day, 12,000 tokens/min; limits are per-organization and can be raised via the Developer plan.

Free modelsLlama 3.1 8B Llama 3.3 70B Qwen 系列 Whisper Large v3 约 17 个免费模型

See console.groq.com's free quota / risk review →console.groq.com

Commercial aggregatorUp

openrouter.ai

Source · cheahjs/free-llm-api-resources

OverviewOpenRouter is a leading commercial LLM aggregation gateway (300+ models, one OpenAI-compatible API); its core business is paid pay-as-you-go routing, with an attached free tier of ~26 :free models.

Free quota:free models are free and require no credit card; rate-limited to ~20 requests/minute, with 50 :free requests/day if you hold <10 credits, rising to 1000/day once you buy >=10 credits (failed requests also count).

Free modelsmeta-llama/llama-3.3-70b-instruct:free qwen/qwen3-coder:free z-ai/glm-4.5-air（:free 层）google/gemma-4-31b-it:free nvidia/nemotron-3-ultra-550b-a55b:free openai/gpt-oss-120b:free+3

See openrouter.ai's free quota / risk review →openrouter.ai

Official (free tier)Up

console.mistral.ai

Source · cheahjs/free-llm-api-resources

OverviewMistral AI's official developer console (La Plateforme) offers an OpenAI-compatible API; its Experiment free tier provides substantial free quota for individual developers, requiring phone verification and opting into data-for-training.

Free quotaExperiment free tier: ~1 req/sec, 500,000 tokens/min, 1,000,000,000 tokens/month; requires phone verification and opting into data-for-training.

Free modelsMistral Large Mistral Small Mistral Nemo Codestral Pixtral Mistral Embed+1

See console.mistral.ai's free quota / risk review →console.mistral.ai

Official (free tier)Up

developers.cloudflare.com

Source · cheahjs/free-llm-api-resources

OverviewCloudflare Workers AI is Cloudflare's edge inference service offering an OpenAI-compatible API with a 10,000 Neurons/day free allowance across ~26 open models.

Free quotaBoth Free and Paid Workers plans get 10,000 Neurons/day at no charge; Free-plan requests fail after the quota is exhausted, Paid overage billed at $0.011 / 1,000 Neurons.

Free modelsGemma 系列 Llama 系列 Qwen 系列 Mistral Llama Guard 3 DeepSeek R1 Distill+1

See developers.cloudflare.com's free quota / risk review →developers.cloudflare.com

Official (free tier)Up

aws.amazon.com

Source · 12britz/awesome-free-models

OverviewAWS (Amazon Bedrock) is Amazon's managed LLM service giving a unified API to Claude, Llama, Titan and more; it is pay-as-you-go with no permanent free tier, only a one-time cloud credit for trials.

Free quotaBedrock has no permanent free tier and bills per use from the first call; accounts opened after 2025-07-15 can receive $200 in AWS credits (expiring in 6 months) usable across services including Bedrock.

Free modelsClaude 系列 Llama 3 Amazon Titan 以及 Bedrock 目录内其他模型

See aws.amazon.com's free quota / risk review →aws.amazon.com

Official (free tier)Up