Table of Free AI API Key Providers for n8n
Website Name Link Notes
Hugging Face Hugging Face Free tier with per-model rate limits; supports
thousands of AI models.
Google AI (Gemini Google AI Free tier with 60 requests per minute; ideal for
API) n8n development.
OpenRouter OpenRouter Free tier with 20 requests per minute; accesses
multiple AI models.
Replicate Replicate $5 free credits for new users; expires after 14
days; good for testing.
Ollama Ollama Free, locally hosted; unlimited usage but requires
hardware setup.
Groq Groq Free tier with 1,000 requests per day; supports
fast inference.
Scaleway Scaleway Free beta until Mar 2025; 100 requests per
Generative Generative minute.
OVH AI Endpoints OVH AI Endpoints Free beta; 12 requests per minute; supports
multiple models.
Cohere Cohere Free tier with 1,000 tokens per month; suitable
for text tasks.
Fireworks AI Fireworks AI $1 free credits; supports 2.5 billion tokens for
testing.
Key Notes for n8n Integration
Hugging Face: Offers a free inference API with rate limits per model. It’s widely compatible with
n8n via HTTP Request nodes or custom integrations for tasks like text generation or embeddings.
Google AI (Gemini API): Provides a generous free tier (60 requests/minute), making it suitable for
n8n workflows involving Google’s Gemini models. Requires an API key, easily obtainable after
signing up.
OpenRouter: Supports multiple models (e.g., Llama, Mistral) with a free tier of 20 requests per
minute. Its unified API simplifies n8n integration for diverse AI tasks.
Replicate: Offers $5 in free credits (expiring after 14 days), useful for testing models in n8n before
committing to paid plans.
Ollama: Completely free and unlimited, but requires local hosting (e.g., on your computer or
server). Use n8n’s HTTP Request node to interact with Ollama’s local API, ideal for cost-free setups
if you have the hardware.
Groq: Free tier with 1,000 requests per day, optimized for fast inference. Easy to integrate into n8n
for tasks like text generation.
Scaleway Generative: Free beta until March 2025 with 100 requests per minute, making it a
strong temporary option for n8n users.
OVH AI Endpoints: Free beta with 12 requests per minute, supporting various models. Compatible
with n8n for lightweight workflows.
Cohere: Free tier with 1,000 tokens per month, suitable for text classification or embeddings in
n8n.
Fireworks AI: $1 in free credits, supporting extensive token usage for testing in n8n workflows.
Important Considerations
No Truly Unlimited Free APIs: Most services impose limits (e.g., requests per minute, tokens per
month, or expiring credits). Ollama is the exception, offering unlimited usage but requiring local
hosting, which may need computational resources (e.g., GPU for large models).
n8n Compatibility: All listed services can be integrated into n8n using HTTP Request nodes or, in
some cases, dedicated nodes (e.g., Hugging Face, Cohere). Check n8n’s community nodes or
documentation for specific integrations.
Exclusions: Services like Puter.js were excluded because, while developers can use it for free, it
operates on a “User Pays” model, shifting costs to end-users, which doesn’t align with “free of cost”
for all parties. Similarly, OpenAI’s API was excluded as it requires payment beyond limited trial
credits.
Verification: Always verify the current status of free tiers, as terms may change (e.g., Scaleway’s
beta ending in March 2025). Check each platform’s documentation for n8n setup instructions.
Why This List?
The list focuses on services explicitly offering free API keys or access, avoiding paid-only platforms like
OpenAI (beyond trial credits) or aimlapi.com, which requires flat pricing. Sources like Hugging Face’s
blog, Analytics Vidhya, and developer tutorials were cross-referenced to ensure accuracy. Platforms
like GitHub repositories sharing “free” OpenAI keys were avoided due to ethical concerns and potential
violations of terms of service.
Recommendations for n8n Users
For Testing: Use Replicate or Fireworks AI for short-term testing with free credits.
For Long-Term Free Use: Consider Ollama if you can host locally, or Google AI/OpenRouter for
generous free tiers with minimal setup.
Check n8n Setup: Ensure you configure API keys in n8n’s Credentials section and test rate limits
to avoid workflow interruptions.
