MiniMax M2.7 vs Qwen3.6 27B
MiniMax M2.7 scores higher on the intelligence index, 23 against 22. The cheapest machine here that runs MiniMax M2.7 is the Mac Studio M5 Ultra, 256GB, at $10,799. Qwen3.6 27B runs on the Framework Desktop, 32GB at $1,269, $9,530 less. On the Mac Studio M5 Ultra, 256GB, the cheapest machine here that runs both, Qwen3.6 27B is about 1.9× quicker: 47 tok/s against 25, both estimated from memory bandwidth. At 500k tokens a day the Mac Studio M5 Ultra, 256GB pays for itself in 173 years running Qwen3.6 27B, against 272 years running MiniMax M2.7.
| MiniMax M2.7 | Qwen3.6 27B | |
|---|---|---|
| Intelligence index | 23 | 22 |
| Class | Haiku-class | Haiku-class |
| Weights | 140 GB | 17 GB |
| Needs at 32k | 149 GB | 19 GB |
| Quantisation | Q4_K_M | Q4_K_M |
| Parameters | 228.7B (10B active) | 27.8B |
| Max context | 200k | 256k |
| API price per 1M | $0.21 in / $0.84 out | $0.25 in / $2 out |
| Licence | other (see repo) | Apache 2.0 |
| Machines here that run it | 1 of 37 | 30 of 37 |
| Cheapest machine that runs it | Mac Studio M5 Ultra, 256GB $10,799 | Strix Halo Framework Desktop, 32GB $1,269 |
Run MiniMax M2.7 on the Mac Studio M5 Ultra, 256GB · or Qwen3.6 27B on the Framework Desktop, 32GB
Jump to: Side by side on the Mac Studio M5 Ultra, 256GB · How much use it takes to pay for the machine · Machines that run one and not the other · The assumptions behind both columns
Side by side on the Mac Studio M5 Ultra, 256GB
The table above gives each model the cheapest machine that runs it, and those are two different machines, so nothing in it is a like-for-like race. The Mac Studio M5 Ultra, 256GB is the cheapest machine here that runs both, so this is the pair doing the same work on the same hardware.
| MiniMax M2.7 | Qwen3.6 27B | |
|---|---|---|
| Speed at 32k | 25 tok/s estimated | 47 tok/s estimated |
| Pay-back on this machine | Pays back in 272 years | Pays back in 173 years |
| API cost per month | $3.80 | $5.47 |
Run MiniMax M2.7 on the Mac Studio M5 Ultra, 256GB · or Qwen3.6 27B
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 Studio M5 Ultra, 256GB. Qwen3.6 27B 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 use | MiniMax M2.7 | Qwen3.6 27B |
|---|---|---|
| 50ka few chats a day | 2,720 years | 1,726 years |
| 200klight assistant use | 680 years | 432 years |
| 1Ma moderate coding-assistant day | 136 years | 86 years |
| 4Mheavy coding with an agent | 34 years | 22 years |
| 20Magents running most of the day | 6.8 years | 4.3 years |
Run MiniMax M2.7 at 20M tokens a day · or Qwen3.6 27B
Machines that run one and not the other
Qwen3.6 27B needs 19 GB of memory at 32k of context and MiniMax M2.7 needs 149 GB. That puts Qwen3.6 27B on 29 of the 37 machines priced here that MiniMax M2.7 does not, starting at $1,269.
| Machine | Price | Memory | Speed on Qwen3.6 27B | Pay-back |
|---|---|---|---|---|
| Strix Halo Framework Desktop, 32GB | $1,269 | 32 GB | 10 tok/s estimated | Pays back in 22 years |
| Mac mini M6, 32GB | $1,299 | 32 GB | 6.7 tok/s estimated | Pays back in 21 years |
| AMD Radeon AI PRO R9700, 32GB | $1,299card only | 32 GB | 29 tok/s estimated | Pays back in 22 years |
| Strix Halo Corsair AI Workstation 300, 64GB | $1,700 | 64 GB | 10 tok/s estimated | Pays back in 29 years |
| Strix Halo Framework Desktop, 64GB | $1,959 | 64 GB | 10 tok/s estimated | Pays back in 33 years |
| NVIDIA GeForce RTX 5090, 32GB | $1,999card only | 32 GB | 65 tok/s estimated | Pays back in 33 years |
23 more, on the Qwen3.6 27B page.
MiniMax M2.7 is also head to head with Ling 3.0 flash above it on the leaderboard. Two more models need much the same memory: DeepSeek V4-Flash and Qwen3 235B-A22B Instruct 2507. Qwen3.6 27B is also head to head with Nemotron 3.5 Bolt 74B-A9B below it on the leaderboard. One more model needs much the same memory: Gemma 3 27B it.
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 MiniMax M2.7 · every machine that runs Qwen3.6 27B · every other match-up · both against the frontier · the quickest pay-back at each level of use