Is Granite 4.0 Micro or GPT-5.3 Codex cheaper?

Granite 4.0 Micro is cheaper on output tokens ($0.11 vs $14.00 per 1M).

Which has the larger context window, Granite 4.0 Micro or GPT-5.3 Codex?

GPT-5.3 Codex has the larger context window (400K tokens).

Granite 4.0 Micro vs GPT-5.3 Codex

Granite 4.0 Micro is cheaper on output tokens, while GPT-5.3 Codex offers a larger context window. Choose Granite 4.0 Micro or GPT-5.3 Codex based on the trade-off between cost, context, and the benchmarks that matter for your use case.

Spec	Granite 4.0 Micro	GPT-5.3 Codex
Provider	OpenRouter	OpenRouter
Input / 1M tokens	$0.02	$1.75
Output / 1M tokens	$0.11	$14.00
Context window	131K	400K
Parameters	—	—
Open weights	Yes	No
Released	Oct 2025	Feb 2026

Granite 4.0 Micro 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.