Home / Compare / MiniMax M3 vs gpt-oss-20b

MiniMax M3 vs gpt-oss-20b

For a typical chat reply (2,000 tokens in, 400 out), MiniMax M3 costs 15× as much as gpt-oss-20b: $1.08 against $0.072 per 1,000 requests at list price on 2026-09-28.

MiniMax M3gpt-oss-20b
Input, per 1M tokens$0.30$0.018
Output, per 1M tokens$1.20$0.09
Context window1.05M131K
Classification or routing (600 tokens in, 10 out), per 1,000$0.192$0.0117
Chat reply (2,000 in, 400 out), per 1,000$1.08$0.072
Retrieval-augmented answer (8,000 in, 500 out), per 1,000$3.00$0.189
Confidence signalyesyes
Tool callingyesyes
BANKING77, correct of 400316313

Which one should carry your traffic?

If gpt-oss-20b answers your requests as well as MiniMax M3, moving that traffic saves 93% on chat-shaped requests. Averages cannot tell you whether it does for your requests. gpt-oss-20b exposes a confidence signal, so Gimbix can let it answer first and hand unsure cases to MiniMax M3.

MiniMax M3 profile · gpt-oss-20b profile · Measure it on your traffic