DeepSeek V3.2 vs GLM 5

DeepSeek V3.2 and GLM 5 match in llmrun Score, with values of 110.4 versus 111.4; neither fits 24 GB.

Model A · DeepSeek

DeepSeek V3.2

110.4llmrun Score

685.4B · 41B active · 160K context · MIT · Dec 2025

Chat

Model B · Z.ai

GLM 5

111.4llmrun Score

753.9B · 42.4B active · 198K context · MIT · Feb 2026

ChatFunctions

Which should you choose?

Choose GLM 5 if you want…

Stronger at agents
117.9 vs 108.1
Ahead on the benchmarks
7 of 8 shared boards
Also does function-calling
DeepSeek V3.2 doesn't list it

DeepSeek V3.2: no clear edge over GLM 5 in the data we have.

On your hardware

Each model at its default quant (Q4_K_M), a smaller one when it must. Estimates, ±20%.

MemoryDeepSeek V3.2GLM 5

8 GB

RTX 4060 · 8 GB tier

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

12 GB

RTX 3060 12GB · 12 GB tier

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

16 GB

RTX 5060 Ti 16GB · 16 GB tier

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

24 GBreference

RTX 4090 · 24 GB tier

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

32 GB

RTX 5090 · 32 GB tier

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

48 GB

RTX A6000 · 48 GB tier

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

Mac 32 GB

Mac Mini M4 · Macs

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

Mac 64 GB

Mac Studio M4 Max · Macs

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

Mac 128 GB

Mac Studio M4 Max · Macs

DeepSeek V3.2

Doesn't fit

smallest quant needs 189.9 GB

GLM 5

Doesn't fit

smallest quant needs 209.1 GB

Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).

llmrun Score

with its 90% range · bars share one scale

DeepSeek V3.2 GLM 5

Overall

DeepSeek V3.2

110.4108–113

GLM 5

111.4109–114

Coding

DeepSeek V3.2

no data

GLM 5

110.2103–118

Agents

DeepSeek V3.2

108.1103–113

GLM 5

117.9112–124

Reasoning

DeepSeek V3.2

112.0108–116

GLM 5

109.1105–113

DeepSeek V3.2 · #24 of 92 open models · #21 of 88 that fit ≤512 GB

GLM 5 · #21 of 92 open models · #18 of 88 that fit ≤512 GB

Benchmarks

8 shared boards: DeepSeek V3.2 leads 1, GLM 5 leads 7 · Benchmarks imported Oct 2026

BoardDeepSeek V3.2GLM 5

SWE-bench Verified

DeepSeek V3.2

70.0%#5/20 open

IndependentSWE-bench · Feb 2026

GLM 5

72.8%#2/20 open

IndependentSWE-bench · Feb 2026

LMArena Text

DeepSeek V3.2

1424.8#29/186 open

ArenaLMArena · Oct 2026

GLM 5

1457.5#12/186 open

ArenaLMArena · Oct 2026

Terminal-Bench

DeepSeek V3.2

39.6%#5/16 open

Lab-reportedEpoch AI · Dec 2025

GLM 5

52.4%#1/16 open

Lab-reportedEpoch AI · Feb 2026

LMArena WebDev

DeepSeek V3.2

1362.2#31/50 open

ArenaLMArena · Oct 2026

GLM 5

1434.4#23/50 open

ArenaLMArena · Oct 2026

ARC-AGI-2

DeepSeek V3.2

4.0%#12/18 open

IndependentEpoch AI · Dec 2025

GLM 5

4.9%#11/18 open

IndependentEpoch AI · Feb 2026

MMLU-Pro

DeepSeek V3.2

85.0%#8/144 open

IndependentTIGER-Lab

GLM 5

86.0%#6/144 open

IndependentTIGER-Lab

SWE-bench Multilingual

DeepSeek V3.2

59.0%#4/4 open

IndependentSWE-bench · Feb 2026

GLM 5

69.7%#1/4 open

IndependentSWE-bench · Feb 2026

ARC-AGI

DeepSeek V3.2

57.0%#11/18 open

IndependentEpoch AI · Dec 2025

GLM 5

44.7%#12/18 open

IndependentEpoch AI · Feb 2026

16 more boards (other shared and single-model boards)

CL-bench

DeepSeek V3.2

12.4%#6/7 open

Lab-reportedEpoch AI · Dec 2025

GLM 5

18.7%#2/7 open

Lab-reportedEpoch AI · Feb 2026

Only DeepSeek V3.2 (7)

Only GLM 5 (8)

VRAM by quantization

Memory to load the model and 2K context; below it, the total at 32K and 128K context. Lighter side in bold.

QuantDeepSeek V3.2GLM 5

Q2_K

DeepSeek V3.2

295 GB245.4 GB file

32K 294 GB · 128K 301 GB

GLM 5

325 GB276.0 GB file

32K 324 GB · 128K 332 GB

Q3_K_S

DeepSeek V3.2

304 GB290.4 GB file

32K 302 GB · 128K 309 GB

GLM 5

334 GB326.3 GB file

32K 333 GB · 128K 342 GB

Q3_K_M

DeepSeek V3.2

338 GB320.2 GB file

32K 337 GB · 128K 344 GB

GLM 5

372 GB360.3 GB file

32K 371 GB · 128K 380 GB

Q4_0

DeepSeek V3.2

347 GB380.1 GB file

32K 345 GB · 128K 352 GB

GLM 5

381 GB427.1 GB file

32K 380 GB · 128K 389 GB

Q4_K_M

DeepSeek V3.2

415 GB405.4 GB file

32K 414 GB · 128K 421 GB

GLM 5

457 GB455.9 GB file

32K 456 GB · 128K 464 GB

Q5_K_M

DeepSeek V3.2

492 GB476.2 GB file

32K 491 GB · 128K 498 GB

GLM 5

541 GB535.2 GB file

32K 540 GB · 128K 549 GB

Q6_K

DeepSeek V3.2

569 GB551.3 GB file

32K 568 GB · 128K 575 GB

GLM 5

626 GB619.4 GB file

32K 625 GB · 128K 634 GB

Q8_0

DeepSeek V3.2

689 GB713.3 GB file

32K 688 GB · 128K 695 GB

GLM 5

758 GB801.3 GB file

32K 757 GB · 128K 766 GB

9 more quants

IQ2_XXS

DeepSeek V3.2

192 GB217.1 GB file

32K 191 GB · 128K 198 GB

GLM 5

212 GB241.1 GB file

32K 211 GB · 128K 219 GB

IQ2_M

DeepSeek V3.2

235 GB228.1 GB file

32K 234 GB · 128K 241 GB

GLM 5

259 GB254.8 GB file

32K 258 GB · 128K 267 GB

IQ3_XXS

DeepSeek V3.2

269 GB273.1 GB file

32K 268 GB · 128K 275 GB

GLM 5

296 GB304.5 GB file

32K 295 GB · 128K 304 GB

IQ4_XS

DeepSeek V3.2

372 GB358.3 GB file

32K 371 GB · 128K 378 GB

GLM 5

409 GB402.9 GB file

32K 408 GB · 128K 417 GB

IQ4_NL

DeepSeek V3.2

389 GB379.1 GB file

32K 388 GB · 128K 395 GB

GLM 5

428 GB425.9 GB file

32K 427 GB · 128K 436 GB

Q4_1

DeepSeek V3.2

389 GB420.8 GB file

32K 388 GB · 128K 395 GB

GLM 5

428 GB472.7 GB file

32K 427 GB · 128K 436 GB

Q4_K_S

DeepSeek V3.2

389 GB381.1 GB file

32K 388 GB · 128K 395 GB

GLM 5

428 GB428.4 GB file

32K 427 GB · 128K 436 GB

Q5_K_S

DeepSeek V3.2

475 GB462.6 GB file

32K 474 GB · 128K 481 GB

GLM 5

523 GB519.7 GB file

32K 522 GB · 128K 530 GB

BF16

DeepSeek V3.2

1375 GB1342.3 GB file

32K 1373 GB · 128K 1380 GB

GLM 5

1512 GB1508.0 GB file

32K 1511 GB · 128K 1520 GB

How to run each

DeepSeek V3.2

Ollama · Q4_K_M

$ollama run hf.co/unsloth/DeepSeek-V3.2-GGUF:Q4_K_M

All ways to run DeepSeek V3.2 →

GLM 5

Ollama · Q4_K_M

$ollama run hf.co/unsloth/GLM-5-GGUF:Q4_K_M

All ways to run GLM 5 →

Specifications

Specifications of DeepSeek V3.2 and GLM 5
Spec	DeepSeek V3.2	GLM 5
Publisher	DeepSeek	Z.ai
Parameters	685.4B	753.9B
Active parameters	41B per token	42B per token
Architecture	DeepseekV32ForCausalLM	GlmMoeDsaForCausalLM
Context	160K tokens	198K tokens
Modalities	Chat	Chat, Function-calling
License	MIT	MIT
Released	Dec 2025	Feb 2026

Frequently Asked Questions

Which is better, DeepSeek V3.2 or GLM 5?

DeepSeek V3.2 and GLM 5 match in llmrun Score, with values of 110.4 versus 111.4; neither fits 24 GB. Choose GLM 5 for stronger at agents and ahead on the benchmarks.

Which needs less VRAM, DeepSeek V3.2 or GLM 5?

At Q4_K_M, DeepSeek V3.2 requires 415.1 GB while GLM 5 needs 456.5 GB, making DeepSeek V3.2 the easier model to run on your local setup.

Can I run DeepSeek V3.2 or GLM 5 on a 16 GB GPU?

DeepSeek V3.2 does not fit (smallest quant needs 189.9 GB); GLM 5 does not fit (smallest quant needs 209.1 GB).

Which has a longer context window, DeepSeek V3.2 or GLM 5?

DeepSeek V3.2 supports 163,840 tokens and GLM 5 supports 202,752 tokens.