AI Software & Agents
Large Language Models
Large language models (LLMs) are the foundation-model layer underneath nearly every major AI lab's products — transformer-based systems trained on massive, increasingly multimodal corpora to generate and reason over language. The 2026 landscape is defined by intense frontier competition (OpenAI's GPT-5, Anthropic's Claude, Google DeepMind's Gemini 3, xAI's Grok 4) alongside a rapidly closing gap from open-weight Chinese models like DeepSeek V4 Pro, which cut its own pricing roughly 75% while staying best-in-class for coding.
Sources
Every claim traced to a primary source — evidence, recent activity, and insider filing behavior, all in one place.
Evidence
2 sources · 57% avg. source confidence
- llm-stats.com — 2026 LLM landscape: GPT-5 family, Claude Opus/Sonnet, Gemini 3 Pro, Grok 4 lead frontier competition58% source confidence
- aimlapi.com — DeepSeek V4 Pro cuts prices ~75% while topping price-to-performance on coding56% source confidence
Dependency Map
What Large Language Models depends on below, and who depends on Large Language Models above — click any node to make it the new center, 3 levels deep.
Connection type
Click any node to make it the new center. Scroll to zoom, drag to pan.
The Story So Far (last 6 months)
Auto-generated from this entity's dated milestones, relationship updates, and sourced evidence — not AI-written, just sorted.
Large Language Models's Timeline
A sourced, dated history of Large Language Models's key moments — founding to present.
Jan 2017 · Transformer architecture introduced
The transformer architecture was introduced, becoming the foundational design underlying nearly every major large language model since.
Jan 2026 · Frontier competition intensifies, open-weight models close the gap
The 2026 landscape is defined by intense frontier competition (GPT-5, Claude Opus/Sonnet, Gemini 3 Pro, Grok 4) alongside a rapidly closing gap from open-weight Chinese models like DeepSeek V4 Pro, which cut its own pricing roughly 75% while remaining best-in-class on price-to-performance for coding.
How These Connections Evolved
How a relationship changed over time, not just its current state — sourced, dated, and traceable.
Large Language Models ↔ OpenAI
Jan 2026
OpenAI's GPT-5 family led frontier proprietary model competition in 2026, as the performance gap versus top open-weight models (DeepSeek V4 Pro, Qwen 3.7 Max) narrowed to single-digit percentages on most benchmarks.
Now
OpenAI develops the GPT frontier model family.
Large Language Models ↔ Anthropic
Jan 2026
Anthropic's Claude Opus and Sonnet models competed at the frontier of the 2026 LLM landscape alongside OpenAI's GPT-5 family and Google's Gemini 3 Pro.
Now
Anthropic develops the Claude family of large language models.
Large Language Models ↔ Alphabet (Google)
Jan 2026
Google DeepMind's Gemini models expanded beyond Google's own products when Apple began powering an updated Siri with Gemini, while Gemini 3 Pro competed at the frontier of the broader 2026 LLM landscape.
Now
Google DeepMind develops the Gemini family of large language models.
Large Language Models ↔ AI Accelerators
2024–2026
LLM training and inference demand drove combined hyperscaler AI data-center capex from roughly $226B (2024) to $410B (2025) toward a projected $725B (2026), with HBM and advanced-packaging supply repeatedly gating available accelerator capacity.
Now
Training and serving large language models depends on AI accelerator hardware at massive scale.