Is Grok 4.20 (Non-Reasoning) or GPT-5.3 Codex cheaper?

Grok 4.20 (Non-Reasoning) is cheaper on output tokens ($6.00 vs $14.00 per 1M).

Which has the larger context window, Grok 4.20 (Non-Reasoning) or GPT-5.3 Codex?

Grok 4.20 (Non-Reasoning) has the larger context window (2M tokens).

Grok 4.20 (Non-Reasoning) vs GPT-5.3 Codex

Grok 4.20 (Non-Reasoning) is cheaper on output tokens, while Grok 4.20 (Non-Reasoning) offers a larger context window. Choose Grok 4.20 (Non-Reasoning) or GPT-5.3 Codex based on the trade-off between cost, context, and the benchmarks that matter for your use case.

Spec	Grok 4.20 (Non-Reasoning)	GPT-5.3 Codex
Provider	LLM Gateway	LLM Gateway
Input / 1M tokens	$2.00	$1.75
Output / 1M tokens	$6.00	$14.00
Context window	2M	400K
Parameters	—	—
Open weights	No	No
Released	Mar 2026	Feb 2026

Grok 4.20 (Non-Reasoning) details →GPT-5.3 Codex details →

FAQ

Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.