GROUNDING

Text / Language Models (page 4 of 6)

2 min read

Text / Language Models

This is the largest and most familiar model category: general-purpose text models trained to follow instructions and hold a conversation — the GPT, Claude, Gemini, Llama, and Qwen families and their many variants. Most day-to-day AI-builder work runs through a model in this category, and the competitive dynamic between labs (context window size, cost per token, instruction-following quality) plays out here first before spreading to more specialized categories below. GROUNDING tracks new releases, version bumps, and benchmark results across this category since it’s the one most builders check first when picking a model for a new project.

At a glance

  • 339 tracked models (63 hand-curated)
  • Most recently updated: Opus 5 (2026-07-25)

Most actively covered

FAQ

What is the Text / Language Models category?

General-purpose chat and instruction-following language models.

How many models are in this category?

GROUNDING currently tracks 339 models in Text / Language Models, 63 of which have a hand-curated profile.

How current is this hub?

The most recently updated entry is Opus 5, last mentioned 2026-07-25; the radar refreshes hourly.

339 pages with this tag.

Incoming · private radar

A radar tuned to your space.

Private, source-tuned weekly digests of the signals and knowledge gaps that matter to you — every claim human-reviewed before it ships.

Request a custom radar
$300–500/mo · 1–2 design partners

Graph View