Cada termo IA · Explicado
Cada benchmark, chip, tipo de memória, conceito, empresa e família de modelos · com dados ao vivo do BenchGecko sob cada definição. O glossário definitivo de IA.
Trilhas de aprendizado
Sequências curadas que contam uma história. Cada trilha percorre 6 a 8 termos em ordem.
The AI Bubble Explained
Seven terms that decode whether AI is overpriced, fairly priced, or criminally underpriced. Read in order.
Pick an AI Model
Six terms to go from "I need an AI" to "here is the cheapest model that meets my spec."
From Sand to Model
The AI supply chain in 7 terms · foundry, memory, chip, system, datacenter, provider, API.
What is GenAI · for normies
Zero jargon. Six terms to sound smart at a dinner party.
The Reasoning Era
Why the second half of 2025 changed pricing forever.
Why AI is Expensive
Six terms for the angry question at the VC meeting.
13 categorias
Benchmarks
47How models are measured · SWE-bench, GPQA, MMLU.
Chips
25Every AI accelerator · H100, B200, MI300X, TPU, Trainium.
Memory
13HBM, DDR, NAND · the bandwidth behind every model.
Model families
7GPT, Claude, Gemini, Llama, Qwen, DeepSeek.
Companies
70Labs, clouds, and startups building the stack.
Foundry
4TSMC, Samsung, Intel · where silicon gets made.
Systems
10DGX, NVL72, Colossus, Aurora · rack-scale builds.
Concepts
37MoE, RAG, reasoning, quantization, tokens, agents.
Pricing
16Input, output, cache, arbitrage, free tiers.
Economy
15Market cap, P/S, bubble index, funding rounds.
Agents
32Claude Code, Codex, Cursor, Cline, MCP.
Compliance
6SOC 2, FedRAMP, GDPR, data residency.
People
15Researchers and founders shaping AI.
Termos novos
OpenAI's Generative Pre-trained Transformer lineage · GPT-1 through GPT-5 and onward.
Anthropic's foundation-model family · Opus / Sonnet / Haiku tiers plus Extended Thinking variants.
Google DeepMind's foundation-model family · multimodal from day one, long context leader.
Meta's open-weight foundation-model family · the dominant open-source LLM lineage.
Alibaba's open-weight model family · dominant in Chinese + multilingual benchmarks.
Hangzhou-based lab behind V3 and R1 · the MoE + RL-reasoning recipe that crashed open-weight prices.
Paris-based AI lab · open-weight (Mistral, Mixtral) and closed (Large) tiers, EU data sovereignty focus.
ARC AI2 is a knowledge benchmark tracked by BenchGecko across every frontier and open-weight model.