AI services with generous free tiers for testing and low-volume usage.
LLM Free Tiers
| Vendor | Free Tier | Limits | Best For |
|---|
| Google Gemini | 1M tokens/mo | RPM limits | General testing |
| OpenAI | 100k tokens/mo | GPT-4o-mini only, 3 RPM | Initial exploration |
| Mistral | 100k tokens/mo | Le Chat | Evaluation |
| Cohere | 10k tokens/mo | Command R | RAG testing |
| Hugging Face | Unlimited | Rate limited | Open models |
| Groq | 14.4k tokens/min | Fast inference | Prototyping |
Free Image AI
| Vendor | Free Tier | Limits |
|---|
| Stable Diffusion | Unlimited | Queue time |
| Hugging Face Spaces | Limited | Shared GPU |
| Microsoft Designer | Limited | 5 images/day |
| Leonardo.ai | 150 tokens | Daily |
Free Audio AI
| Vendor | Free Tier | Limits |
|---|
| Whisper API | $0 free credit | $5 initial credit |
| AssemblyAI | 5 hours | Limited models |
| ElevenLabs | 10k chars/mo | Limited voices |
| Google TTS | 4M chars/mo | Standard voices |
Free Embeddings
| Vendor | Free Tier | Limits |
|---|
| Hugging Face Instructor | Unlimited | Local compute |
| Cohere | 10k tokens/mo | Limited models |
| Groq | 14.4k tokens/min | Limited models |
Recommended Free Stack
| Purpose | Service | Monthly Limit |
|---|
| Chatbot prototype | Gemini 1.5 Flash | 1M tokens |
| Image generation | Stable Diffusion | Unlimited |
| Speech transcription | Whisper API | $5 credit |
| Embeddings for RAG | Instructor | Unlimited |
| Code generation | Groq Mixtral | 14.4k/min |
Limitations
- Free tiers typically have rate limits
- May not include latest models
- Commercial use often restricted
- No SLA guarantees