Is Magistral Small 2506 or GPT-OSS 20B cheaper?

GPT-OSS 20B is cheaper on output tokens ($0.14 vs $1.50 per 1M).

Which has the larger context window, Magistral Small 2506 or GPT-OSS 20B?

Magistral Small 2506 has the larger context window (128K tokens).

Magistral Small 2506 vs GPT-OSS 20B

GPT-OSS 20B is cheaper on output tokens, while Magistral Small 2506 offers a larger context window. Choose Magistral Small 2506 or GPT-OSS 20B based on the trade-off between cost, context, and the benchmarks that matter for your use case.

Spec	Magistral Small 2506	GPT-OSS 20B
Provider	IO.NET	IO.NET
Input / 1M tokens	$0.50	$0.03
Output / 1M tokens	$1.50	$0.14
Context window	128K	64K
Parameters	—	21B
Open weights	No	Yes
Released	Jun 2025	Dec 2024

Magistral Small 2506 details →GPT-OSS 20B details →

FAQ

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