DeepSeek V3.2 vs MiniMax M3
DeepSeek V3.2 and MiniMax M3 match in llmrun Score, with values of 110.4 versus 111.0; neither fits 24 GB.
Model A · DeepSeek
DeepSeek V3.2685.4B · 41B active · 160K context · MIT · Dec 2025
Model B · MiniMax
MiniMax M3427.0B · 19.8B active · 1M context · Other · Jun 2026
Choose DeepSeek V3.2 if you want…
Also does chat
MiniMax M3 doesn't list it
More permissive license
MIT vs Other
Choose MiniMax M3 if you want…
Ahead on the benchmarks
4 of 5 shared boards
Lighter on memory
256.7 vs 415.1 GB at Q4_K_M
Longer context window
1M vs 160K tokens
Also does vision, function-calling
DeepSeek V3.2 doesn't list it
Newer release
Jun 2026 vs Dec 2025
On your hardware
Each model at its default quant (Q4_K_M), a smaller one when it must. Estimates, ±20%.
8 GB
RTX 4060 · 8 GB tier
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
12 GB
RTX 3060 12GB · 12 GB tier
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
16 GB
RTX 5060 Ti 16GB · 16 GB tier
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
24 GBreference
RTX 4090 · 24 GB tier
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
32 GB
RTX 5090 · 32 GB tier
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
48 GB
RTX A6000 · 48 GB tier
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
Mac 32 GB
Mac Mini M4 · Macs
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
Mac 64 GB
Mac Studio M4 Max · Macs
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
Mac 128 GB
Mac Studio M4 Max · Macs
DeepSeek V3.2
Doesn't fit
smallest quant needs 189.9 GB
MiniMax M3
IQ2_XXS · ~38 tok/s
119.2 GB · up to 64Kfits
Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).
Overall
DeepSeek V3.2
MiniMax M3
Coding
DeepSeek V3.2
no data
MiniMax M3
Agents
DeepSeek V3.2
MiniMax M3
Math
DeepSeek V3.2
no data
MiniMax M3
Science
DeepSeek V3.2
no data
MiniMax M3
Reasoning
DeepSeek V3.2
MiniMax M3
DeepSeek V3.2 · #24 of 92 open models · #21 of 88 that fit ≤512 GB
MiniMax M3 · #22 of 92 open models · #19 of 88 that fit ≤512 GB
DeepSeek V3.2
1424.8#29/186 open
ArenaLMArena · Oct 2026
MiniMax M3
1440.5#20/186 open
ArenaLMArena · Oct 2026
DeepSeek V3.2
1362.2#31/50 open
ArenaLMArena · Oct 2026
MiniMax M3
1481.7#16/50 open
ArenaLMArena · Oct 2026
DeepSeek V3.2
85.6%#15/68 open
Lab-reportedEpoch AI · Dec 2025
MiniMax M3
78.9%#24/68 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V3.2
28.8%#25/54 open
Lab-reportedEpoch AI · Dec 2025
MiniMax M3
33.7%#17/54 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V3.2
8.0%#18/26 open
Lab-reportedEpoch AI · Dec 2025
MiniMax M3
18.0%#13/26 open
Lab-reportedEpoch AI · Jun 2026
26 more boards (other shared and single-model boards)
DeepSeek V3.2
21.3%#15/16 open
Lab-reportedEpoch AI · Dec 2025
MiniMax M3
37.7%#10/16 open
Lab-reportedEpoch AI · Jun 2026
Only DeepSeek V3.2 (10)
- SWE-bench Verified70.0%
- Terminal-Bench39.6%
- ARC-AGI-24.0%
- MMLU-Pro85.0%
- SWE-bench Multilingual59.0%
- SWE-bench Bash Only60.0%
- ARC-AGI57.0%
- SWE-bench Lite30.7%
- CL-bench12.4%
- CL-bench Life7.4%
Only MiniMax M3 (15)
- GPQA Diamond90.9%
- LiveBench Coding68.2
- LiveBench Reasoning74.5
- AIME 2024/202571.1%
- LiveBench Math77.0
- SciCode47.1%
- SimpleBench45.8%
- FrontierCode14.7%
- Chess Puzzles14.0%
- CritPt3.7%
- ForecastBench61.4
- Mystery Game Puzzles8.0%
- Vending-Bench 22157.8
- ALE-Bench640.0
- Surface Evolver Bench55.0%
VRAM by quantization
Memory to load the model and 2K context; below it, the total at 32K and 128K context. Lighter side in bold.
Q2_K
DeepSeek V3.2
295 GB245.4 GB file
32K 294 GB · 128K 301 GB
MiniMax M3
182 GB151.9 GB file
32K 185 GB · 128K 194 GB
Q3_K_S
DeepSeek V3.2
304 GB290.4 GB file
32K 302 GB · 128K 309 GB
MiniMax M3
187 GB187.3 GB file
32K 190 GB · 128K 199 GB
Q3_K_M
DeepSeek V3.2
338 GB320.2 GB file
32K 337 GB · 128K 344 GB
MiniMax M3
209 GB195.5 GB file
32K 212 GB · 128K 221 GB
Q4_0
DeepSeek V3.2
347 GB380.1 GB file
32K 345 GB · 128K 352 GB
MiniMax M3
214 GB243.6 GB file
32K 217 GB · 128K 226 GB
Q4_K_M
DeepSeek V3.2
415 GB405.4 GB file
32K 414 GB · 128K 421 GB
MiniMax M3
257 GB264.9 GB file
32K 260 GB · 128K 269 GB
Q5_K_M
DeepSeek V3.2
492 GB476.2 GB file
32K 491 GB · 128K 498 GB
MiniMax M3
305 GB318.7 GB file
32K 308 GB · 128K 317 GB
Q6_K
DeepSeek V3.2
569 GB551.3 GB file
32K 568 GB · 128K 575 GB
MiniMax M3
353 GB355.3 GB file
32K 356 GB · 128K 365 GB
Q8_0
DeepSeek V3.2
689 GB713.3 GB file
32K 688 GB · 128K 695 GB
MiniMax M3
428 GB453.6 GB file
32K 430 GB · 128K 439 GB
15 more quants
IQ2_XXS
DeepSeek V3.2
192 GB217.1 GB file
32K 191 GB · 128K 198 GB
MiniMax M3
118 GB134.9 GB file
32K 121 GB · 128K 130 GB
IQ2_XS
DeepSeek V3.2
not listed
MiniMax M3
129 GB129.5 GB file
32K 131 GB · 128K 140 GB
IQ2_S
DeepSeek V3.2
not listed
MiniMax M3
134 GB132.1 GB file
32K 137 GB · 128K 146 GB
IQ2_M
DeepSeek V3.2
235 GB228.1 GB file
32K 234 GB · 128K 241 GB
MiniMax M3
145 GB135.1 GB file
32K 147 GB · 128K 157 GB
IQ3_XXS
DeepSeek V3.2
269 GB273.1 GB file
32K 268 GB · 128K 275 GB
MiniMax M3
166 GB160.3 GB file
32K 169 GB · 128K 178 GB
IQ3_XS
DeepSeek V3.2
not listed
MiniMax M3
177 GB196.4 GB file
32K 179 GB · 128K 189 GB
IQ3_S
DeepSeek V3.2
not listed
MiniMax M3
182 GB175.6 GB file
32K 185 GB · 128K 194 GB
IQ3_M
DeepSeek V3.2
not listed
MiniMax M3
193 GB205.5 GB file
32K 195 GB · 128K 205 GB
Q3_K_L
DeepSeek V3.2
not listed
MiniMax M3
219 GB205.0 GB file
32K 222 GB · 128K 231 GB
IQ4_XS
DeepSeek V3.2
372 GB358.3 GB file
32K 371 GB · 128K 378 GB
MiniMax M3
230 GB208.5 GB file
32K 233 GB · 128K 242 GB
IQ4_NL
DeepSeek V3.2
389 GB379.1 GB file
32K 388 GB · 128K 395 GB
MiniMax M3
241 GB212.7 GB file
32K 244 GB · 128K 253 GB
Q4_1
DeepSeek V3.2
389 GB420.8 GB file
32K 388 GB · 128K 395 GB
MiniMax M3
241 GB268.9 GB file
32K 244 GB · 128K 253 GB
Q4_K_S
DeepSeek V3.2
389 GB381.1 GB file
32K 388 GB · 128K 395 GB
MiniMax M3
241 GB248.6 GB file
32K 244 GB · 128K 253 GB
Q5_K_S
DeepSeek V3.2
475 GB462.6 GB file
32K 474 GB · 128K 481 GB
MiniMax M3
294 GB299.7 GB file
32K 297 GB · 128K 306 GB
BF16
DeepSeek V3.2
1375 GB1342.3 GB file
32K 1373 GB · 128K 1380 GB
MiniMax M3
855 GB852.9 GB file
32K 857 GB · 128K 866 GB
DeepSeek V3.2
Ollama · Q4_K_M
ollama run hf.co/unsloth/DeepSeek-V3.2-GGUF:Q4_K_MMiniMax M3
Ollama · Q4_K_M
ollama run hf.co/unsloth/MiniMax-M3-GGUF:Q4_K_M| Spec | DeepSeek V3.2 | MiniMax M3 |
|---|---|---|
| Publisher | DeepSeek | MiniMax |
| Parameters | 685.4B | 427.0B |
| Active parameters | 41B per token | 20B per token |
| Architecture | DeepseekV32ForCausalLM | MiniMaxM3SparseForConditionalGeneration |
| Context | 160K tokens | 1M tokens |
| Modalities | Chat | Vision, Function-calling |
| License | MIT | Other |
| Released | Dec 2025 | Jun 2026 |
Frequently Asked Questions
Which is better, DeepSeek V3.2 or MiniMax M3?
DeepSeek V3.2 and MiniMax M3 match in llmrun Score, with values of 110.4 versus 111.0; neither fits 24 GB. Choose DeepSeek V3.2 for also does chat and more permissive license; choose MiniMax M3 for ahead on the benchmarks and lighter on memory.
Which needs less VRAM, DeepSeek V3.2 or MiniMax M3?
At Q4_K_M, DeepSeek V3.2 requires 415.1 GB while MiniMax M3 needs 256.7 GB, making MiniMax M3 the easier model to run on your local setup.
Can I run DeepSeek V3.2 or MiniMax M3 on a 16 GB GPU?
DeepSeek V3.2 does not fit (smallest quant needs 189.9 GB); MiniMax M3 does not fit (smallest quant needs 119.2 GB).
Which has a longer context window, DeepSeek V3.2 or MiniMax M3?
DeepSeek V3.2 supports 163,840 tokens and MiniMax M3 supports 1,048,576 tokens.