In this file, we will collect all interesting links
OpenAI and GPT are not the only hosted LLMs that we can use. There are other services that we can use
- mistral.ai (5€ free credit on sign up)
- Groq (can inference from open source LLMs with rate limits)
- TogetherAI (can inference from variety of open source LLMs, 25$ free credit on sign up)
- Google Gemini (2 months unlimited access)
- OpenRouterAI (some small open-source models, such as Gemma 7B, are free)
- HuggingFace API (over 150,000 open-source models, rate-limited and free)
- Cohere (provides a developer trail key which allows upto 100 reqs/min for generating, summarizing, and classifying text. Read more here)
- wit (Facebook AI Afiliate - free)
- Anthropic API (starting from $0.25 / MTok for input and $1.25 / MTok for the output for the most affordable model)
- AI21Labs API (Free trial including $10 credits for 3 months)
- Replicate (faster inference, can host any ML model. charges 0.10$ per 1M input tokens for llama/Mistral model)
These services help run LLMs locally, also without GPUs
- Text Generation
- Chatbots
- Sentiment Analysis
- Summarization
- Translation
- Guides and Tutorials
- Tools and Frameworks
- Colab Notebooks
- Techniques and Best Practices
- Tools
- Examples and Case Studies
- Hosting Services
- Serverless Deployments
- Containerization
- Best Practices
- Frameworks and Guidelines
- Tools
- Research Papers