Sunk Cost sunkcost.ai Data checked 2026-09-03

gpt-oss-20b vs Ministral 3 14B

gpt-oss-20b scores higher on the intelligence index, 9 against 6. Both take the same machine to start: the cheapest here that runs either is the Mac mini M6, 24GB, at $1,099. On it, gpt-oss-20b is about 1.9× quicker: 18 tok/s against 9.4, both estimated from memory bandwidth. At 500k tokens a day the Mac mini M6, 24GB pays for itself in 34 years running Ministral 3 14B, against 426 years running gpt-oss-20b.

gpt-oss-20bMinistral 3 14B
Intelligence index96
ClassBelow every hosted tierBelow every hosted tier
Weights12 GB8.2 GB
Needs at 32k13 GB14 GB
QuantisationMXFP4Q4_K_M
Parameters20.9B (3.6B active)14B
Max context128k256k
API price per 1M$0.02 in / $0.1 out$0.2 in / $0.2 out
LicenceApache 2.0Apache 2.0
Machines here that run it33 of 3733 of 37
Cheapest machine that runs itMac mini M6, 24GB $1,099Mac mini M6, 24GB $1,099
Summarising good not rated
Translation usable not rated
Everyday coding good not rated
Reasoning & maths usable not rated
Agentic work usable not rated

Run gpt-oss-20b on the Mac mini M6, 24GB · or Ministral 3 14B

Ratings are coarse on purpose: they say what a model is usable for, not where it places to the decimal.

Side by side on the Mac mini M6, 24GB

The cheapest machine that runs either model is the same one, so this is the pair doing the same work on the same hardware: Mac mini M6, 24GB, at $1,099.

gpt-oss-20bMinistral 3 14B
Speed at 32k18 tok/s estimated9.4 tok/s estimated
Pay-back on this machinePays back in 426 yearsPays back in 34 years
API cost per month$0.38$3.04

Run gpt-oss-20b on the Mac mini M6, 24GB · or Ministral 3 14B

How much use it takes to pay for the machine

Everything above is at 500k tokens a day. Pay-back moves with how much you actually run, so here are both models at the five levels of use the calculator names, on the Mac mini M6, 24GB. Ministral 3 14B pays for it sooner at every level of use, so which of them to run does not turn on how hard you work it.

A day's usegpt-oss-20bMinistral 3 14B
50ka few chats a day4,263 years335 years
200klight assistant use1,066 years84 years
1Ma moderate coding-assistant day213 years17 years
4Mheavy coding with an agent53 years4.2 years
20Magents running most of the day11 years16 monthsits ceiling

The Mac mini M6, 24GB generates at most 13M tokens a day on Ministral 3 14B, so that column's figure at 20M tokens a day is for the most it can do, not for the whole of what was asked.

Run gpt-oss-20b at 20M tokens a day · or Ministral 3 14B

Memory is not what separates them

gpt-oss-20b needs 13 GB of memory at 32k of context and Ministral 3 14B needs 14 GB. Every machine priced here that runs one runs the other, at 32k of context. So the choice between them is what each is good at, how fast it runs and what the same work costs on an API, not what you have to buy to hold it.

gpt-oss-20b is also head to head with Qwen3-Coder Next above it on the leaderboard and Gemma 4 E4B below it. Ministral 3 14B is also head to head with Llama 4 Scout 17B-16E above it on the leaderboard and Qwen3 8B below it. One more model needs much the same memory: Qwen3 14B.

The assumptions behind both columns

Both columns use the same usage: 500k tokens a day at 15:1 input to output, 32k of context, $0.17 per kWh, and today's API prices held flat. Speeds marked estimated are worked out from memory bandwidth rather than measured, and pay-back scales with them. Where nobody rents an open model by the token, its API prices are the nearest hosted model's, named beside them. Machines are the 37 here with a published price that are still sold. Change any of it in the calculator.

More head to head: every machine that runs gpt-oss-20b · every machine that runs Ministral 3 14B · every other match-up · both against the frontier · the quickest pay-back at each level of use