AI Tools & Resources

OpenAI's flagship product. Best starting point for understanding what LLMs can do.

Anthropic's assistant. Exceptional at long documents, coding, and nuanced reasoning.

Test Gemini models with full API access. Free generous quota for development.

Run open-weight models (Llama, Mistral, and more) at extraordinary inference speed. Best for latency benchmarking.

AI answer engine that searches the live web and cites sources — best for research questions, not just chat.

Run open-weight models locally on your own machine — no API key, no cloud, full privacy.

Anthropic's agentic coding CLI — plans, edits across files, runs tests, and opens PRs from a terminal.

AI-native code editor built around agentic edits, codebase-wide context, and inline chat.

The original AI pair programmer — now with agent mode for multi-file, multi-step changes.

Agentic IDE with a built-in multi-step "flow" mode for larger autonomous coding tasks.

GPT-4o, o1, embeddings, DALL-E 3 and Whisper — the most widely used AI API.

Claude API with a large context window, strong tool use, vision, and safety controls.

Open-source model hub. Run Llama, Mistral, and thousands of specialised models.

Fast, affordable API for open-source models. Good for cost-sensitive applications.

Managed vector database. Easiest to get started, scales to billions of vectors.

Open-source vector DB with hybrid search (vector + keyword). Self-host or cloud.

Lightweight local vector DB — the fastest way to prototype a RAG pipeline.

High-performance vector engine with rich filtering. Great for on-premise setups.

The most widely used LLM framework. Chains, RAG, agents, and 100+ integrations.

Build stateful agent workflows as graphs. Best for complex multi-step reasoning.

Data framework for LLMs. Specialises in connecting models to documents and databases.

Multi-agent framework with role-based agents. Simple API for collaborative AI teams.

Microsoft Research's multi-agent conversation framework. Good for code generation flows.

Tracing, prompt management and evaluation for LangChain apps. Best-in-class UX.

Open-source LLM observability — traces, evals, prompt versioning, cost analytics.

Evaluation platform for LLM apps. Run evals, track regressions, compare prompts.

ML experiment tracking and model monitoring. Widely used in fine-tuning workflows.

Parameter-efficient fine-tuning library — LoRA, QLoRA, and adapters in one package.

2x faster fine-tuning with 70% less memory. Supports Llama 3, Mistral, Gemma.

Serverless GPU compute. Run fine-tuning jobs with a simple Python decorator.

Managed fine-tuning for OpenAI's smaller models. No GPU setup needed.

OpenAI's image generation models — strong prompt adherence and text rendering inside images.

Open-source image generation. Run locally on a laptop GPU or via API.

OpenAI open-source speech-to-text. Best-in-class transcription, 99 languages.

Realistic text-to-speech and voice cloning. Best quality for production voice AI.

Andrew Ng's short courses on LLMs, RAG, agents and fine-tuning. Practical and fast.

Practical deep learning — top-down, code-first. Best for understanding model internals.

Build GPT from scratch in Python. The best video series for understanding transformers deeply.

The 2017 paper that introduced the Transformer architecture. Essential primary source.

OpenAI researcher's deep technical posts on agents, RAG, RLHF, and diffusion models.