Skip to content

Grok 4.20 Heavy

Released 2026-03 · reasoning · 1.0M tokens · 0 benchmarks

Editorial notes

Multi-agent con 16 specialized agents en paralelo. SuperGrok Heavy tier exclusivo. Entrenado en Colossus (200K GPUs). ~3T parametros (figura citada por terceros, no confirmada por xAI). Pricing solo en SuperGrok Heavy $300/mes (no API directa).

Spec sheet

Company
xAI
Country
US
Type
reasoning
Release
2026-03
Context
1.0M tokens
Slug
grok-4-20-heavy
No scores published for this model. The editorial notes have additional context.
Cite this model
BibTeX · APA

BibTeX

@misc{frontier-grok-4-20-heavy,
  title  = {Grok 4.20 Heavy},
  author = {{xAI}},
  year   = {2026},
  note   = {Frontier Benchmarks AI atlas. Accessed 2026-05-08},
  url    = {https://frontierbenchmarks.com/models/grok-4-20-heavy}
}

APA

xAI (2026). Grok 4.20 Heavy [Large language model]. Frontier Benchmarks AI. Retrieved 2026-05-08, from https://frontierbenchmarks.com/models/grok-4-20-heavy

Citation reflects the atlas page, not the original model paper. For the paper, see the "Resources" section above.