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

Tencent Hy3 vs GLM-5.3-SLIM

Tencent Hy3 scores 26 on the intelligence index, and GLM-5.3-SLIM has not been placed on it. None of the 37 machines here still sold at a published price runs Tencent Hy3 at 32k of context. GLM-5.3-SLIM runs on the Mac Studio M5 Ultra, 256GB, from $10,799. At 500k tokens a day that machine pays for itself in 502 years running GLM-5.3-SLIM. Shorten the context to 16k and that machine holds both.

Tencent Hy3GLM-5.3-SLIM
Intelligence index26not placed
ClassHaiku-classnot yet placed
Weights182 GB180 GB
Needs at 32k193 GB183 GB
QuantisationQ4_K_MUD-IQ2_M
Parameters295B (21B active)563B (40B active)
Max context256k1M
API price per 1M$0.0825 in / $0.33 out$0.0699 in / $1.4 outpriced as GLM-5.3
LicenceApache 2.0GLM-5.3 License
Machines here that run it0 of 371 of 37
Cheapest machine that runs itnone listedMac Studio M5 Ultra, 256GB $10,799

Run Tencent Hy3 on the Mac Studio M5 Ultra, 256GB at 16k of context · or GLM-5.3-SLIM

Jump to: Side by side on the Mac Studio M5 Ultra, 256GB at 16k of context · 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 at 16k of context

Tencent Hy3 needs 193 GB at 32k of context. The biggest machine on this list is the Mac Studio M5 Ultra, 256GB, and it has 192 GB usable. Tencent Hy3's weights are 182 GB of that. The rest is the cache, and the cache is the part that shrinks when you ask for less context: at 16k the model needs 188 GB, which that machine holds. It already runs GLM-5.3-SLIM, so at 16k of context one machine here runs both. It costs $10,799. The rest of this page is what each model does with it.

Tencent Hy3GLM-5.3-SLIM
Speed at 16k20 tok/s estimated25 tok/s estimated
Pay-back on this machinePays back in 1,031 yearsPays back in 488 years
API cost per month$1.49$2.33priced as GLM-5.3

Run Tencent Hy3 on the Mac Studio M5 Ultra, 256GB · or GLM-5.3-SLIM

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 at 16k of context. GLM-5.3-SLIM 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 useTencent Hy3GLM-5.3-SLIM
50ka few chats a day10,305 years4,875 years
200klight assistant use2,576 years1,219 years
1Ma moderate coding-assistant day515 years244 years
4Mheavy coding with an agent129 years61 years
20Magents running most of the day26 years12 years

Run Tencent Hy3 at 20M tokens a day · or GLM-5.3-SLIM

Machines that run one and not the other

GLM-5.3-SLIM needs 183 GB of memory at 32k of context and Tencent Hy3 needs 193 GB. That puts GLM-5.3-SLIM on 1 of the 37 machines priced here that Tencent Hy3 does not, starting at $10,799.

MachinePriceMemorySpeed on GLM-5.3-SLIMPay-back
Mac Studio M5 Ultra, 256GB$10,799256 GB23 tok/s estimatedPays back in 502 years

Tencent Hy3 is also head to head with Inkling Small above it on the leaderboard and Ling 3.0 flash below it. One more model needs much the same memory: GLM-5.3-Flash.

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. The two sections that need one machine to hold both models are at 16k of context instead, which is the longest on the calculator's list where one does. 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: what Tencent Hy3 needs · every machine that runs GLM-5.3-SLIM · every other match-up · both against the frontier · the quickest pay-back at each level of use