Gen-AI Developer Classroom notes 17/Mar/2026

Terms of LLM

  • Temperature: range (0-2)
    • less temperature => factual information
    • more temperature => creative
  • Context length: Number of Tokens (input or output) Give me a list of models with context lengths in descending order, Ensure latest models are also considered
  • Token

LLM Costs

  • We have two types of LLMS
    • Opensource LLM: The cost is the infra i.e. servers with GPUs
    • Propietary LLMS: Cost is generally usage based i.e token based Give me cost of popular Anthropic, OPENAI, Gemini Models
  • LLMS are hosted by Cloud Providers (AWS, Azure, GCP) as Services
    • GCP Vertex AI:
      • Google Propietary Models (Gemini)
      • Claude
      • Opensource LLMS
        • meta
        • qwen
        • deepseek
        • gemma
        • phi
        • mistral
        • gpt-oss
    • AWS Bedrock
      • Claude
      • AWS Titan
      • Opensource LLMS
        • meta
        • qwen
        • deepseek
        • gemma
        • phi
        • mistral
        • gpt-oss
    • Azure OpenAI & Azure AI Foundry
      • OpenAI
      • Claude
      • Grok
      • Opensource LLMS
        • meta
        • qwen
        • deepseek
        • gemma
        • phi
        • mistral
        • gpt-oss

Our Lab Plan

  • GCP => 3 months with 300$ credited (GCP free tier)
  • AWS => 1 year (100$ + 100$)
  • Azure => 1 month (200$)

Prompt Engineering

  • Refer Here for prompts with categories
  • Claude => PromptEnginering Docs

  • Find a prompt which acts as a critique and criticizes your ideas so that we can improve.

By continuous learner

enthusiastic technology learner

Leave a Reply

Discover more from Direct AI Powered By Quality Thought

Subscribe now to keep reading and get access to the full archive.

Continue reading