DeepSeek V4 Flash vs MiniMax M3
DeepSeek V4 Flash leads llmrun Score 115.3 to 112.6 (ranges overlap); neither fits 24 GB.
Run DeepSeek V4 Flash in the cloud →
Model A · DeepSeek
DeepSeek V4 Flash290.9B · 14.1B active · 1M context · MIT · Apr 2026
Model B · MiniMax
MiniMax M3427.0B · 19.8B active · 1M context · Other · Jun 2026
Choose DeepSeek V4 Flash if you want…
Higher llmrun Score
115.3 vs 112.6, ranges overlap
Ahead on the benchmarks
9 of 12 shared boards
Lighter on memory
174.9 vs 256.7 GB at Q4_K_M
Also does chat
MiniMax M3 doesn't list it
More permissive license
MIT vs Other
Choose MiniMax M3 if you want…
Stronger at science
118.0 vs 111.2
Also does vision, function-calling
DeepSeek V4 Flash doesn't list it
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 V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
12 GB
RTX 3060 12GB · 12 GB tier
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
16 GB
RTX 5060 Ti 16GB · 16 GB tier
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
24 GBreference
RTX 4090 · 24 GB tier
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
32 GB
RTX 5090 · 32 GB tier
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
48 GB
RTX A6000 · 48 GB tier
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
Mac 32 GB
Mac Mini M4 · Macs
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
Mac 64 GB
Mac Studio M4 Max · Macs
DeepSeek V4 Flash
Doesn't fit
smallest quant needs 80.5 GB
MiniMax M3
Doesn't fit
smallest quant needs 119.2 GB
Mac 128 GB
Mac Studio M4 Max · Macs
DeepSeek V4 Flash
Q2_K · ~41 tok/s
124.1 GB · up to 32K
MiniMax M3
IQ2_XXS · ~38 tok/s
119.2 GB · up to 64K
Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).
Overall
DeepSeek V4 Flash
MiniMax M3
Coding
DeepSeek V4 Flash
MiniMax M3
Agents
DeepSeek V4 Flash
no data
MiniMax M3
Math
DeepSeek V4 Flash
no data
MiniMax M3
Science
DeepSeek V4 Flash
MiniMax M3
Reasoning
DeepSeek V4 Flash
MiniMax M3
DeepSeek V4 Flash · #24 of 106 open models · #8 of 82 that fit ≤192 GB
MiniMax M3 · #27 of 106 open models · #23 of 102 that fit ≤512 GB
Benchmarks
12 shared boards: DeepSeek V4 Flash leads 9, MiniMax M3 leads 3 · Benchmarks imported Nov 2026
DeepSeek V4 Flash
69.2#18/20 open
IndependentLiveBench · Jun 2026
MiniMax M3
68.2#20/20 open
IndependentLiveBench · Jun 2026
DeepSeek V4 Flash
70.6#19/20 open
IndependentLiveBench · Jun 2026
MiniMax M3
74.5#18/20 open
IndependentLiveBench · Jun 2026
DeepSeek V4 Flash
79.7#18/20 open
IndependentLiveBench · Jun 2026
MiniMax M3
77.0#20/20 open
IndependentLiveBench · Jun 2026
DeepSeek V4 Flash
44.9%#28/72 open
Lab-reportedEpoch AI · Apr 2026
MiniMax M3
47.1%#22/72 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V4 Flash
46.3%#11/25 open
IndependentEpoch AI · Apr 2026
MiniMax M3
45.8%#12/25 open
IndependentEpoch AI · Jun 2026
DeepSeek V4 Flash
678.2#28/48 open
Lab-reportedEpoch AI · Apr 2026
MiniMax M3
640.0#29/48 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V4 Flash
7.1%#19/84 open
Lab-reportedEpoch AI · Apr 2026
MiniMax M3
3.7%#25/84 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V4 Flash
86.4%#15/70 open
Lab-reportedEpoch AI · Apr 2026
MiniMax M3
78.9%#25/70 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V4 Flash
77.0#9/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
70.9#11/19 open
IndependentLiveBench · Oct 2026
DeepSeek V4 Flash
36.6#6/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
12.4#16/19 open
IndependentLiveBench · Oct 2026
20 more boards (other shared and single-model boards)
DeepSeek V4 Flash
70.6#17/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
74.5#15/19 open
IndependentLiveBench · Oct 2026
DeepSeek V4 Flash
35.9%#14/56 open
Lab-reportedEpoch AI · Apr 2026
MiniMax M3
33.7%#18/56 open
Lab-reportedEpoch AI · Jun 2026
DeepSeek V4 Flash
1438.7#21/188 open
ArenaLMArena · Oct 2026
MiniMax M3
1439.2#20/188 open
ArenaLMArena · Oct 2026
DeepSeek V4 Flash
1581.8#14/55 open
ArenaLMArena · Oct 2026
MiniMax M3
1483.3#19/55 open
ArenaLMArena · Oct 2026
DeepSeek V4 Flash
0.0#15/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
8.0#12/19 open
IndependentLiveBench · Oct 2026
DeepSeek V4 Flash
68.0#13/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
76.2#8/19 open
IndependentLiveBench · Oct 2026
DeepSeek V4 Flash
71.7#9/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
66.2#12/19 open
IndependentLiveBench · Oct 2026
DeepSeek V4 Flash
79.2#4/19 open
IndependentLiveBench · Oct 2026
MiniMax M3
68.0#13/19 open
IndependentLiveBench · Oct 2026
Only DeepSeek V4 Flash (1)
- WeirdML45.6%
Only MiniMax M3 (11)
- GPQA Diamond90.9%
- AIME 2024/202571.1%
- FrontierCode14.7%
- APEX-Agents37.7%
- Chess Puzzles17.0%
- ForecastBench64.3
- Mystery Game Puzzles8.0%
- Vending-Bench 22157.8
- Surface Evolver Bench55.0%
- GBA Eval0.9%
- ProofBench18.0%
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 V4 Flash
124 GB107.0 GB file
32K 124 GB · 128K 125 GB
MiniMax M3
182 GB151.9 GB file
32K 185 GB · 128K 194 GB
Q3_K_S
DeepSeek V4 Flash
not listed
MiniMax M3
187 GB187.3 GB file
32K 190 GB · 128K 199 GB
Q3_K_M
DeepSeek V4 Flash
142 GB129.3 GB file
32K 143 GB · 128K 144 GB
MiniMax M3
209 GB195.5 GB file
32K 212 GB · 128K 221 GB
Q4_0
DeepSeek V4 Flash
not listed
MiniMax M3
214 GB243.6 GB file
32K 217 GB · 128K 226 GB
Q4_K_M
DeepSeek V4 Flash
175 GB174.9 GB file
32K 175 GB · 128K 176 GB
MiniMax M3
257 GB264.9 GB file
32K 260 GB · 128K 269 GB
Q5_K_M
DeepSeek V4 Flash
208 GB~207.3 GB file
32K 208 GB · 128K 209 GB
MiniMax M3
305 GB318.7 GB file
32K 308 GB · 128K 317 GB
Q6_K
DeepSeek V4 Flash
240 GB~240.0 GB file
32K 241 GB · 128K 242 GB
MiniMax M3
353 GB355.3 GB file
32K 356 GB · 128K 365 GB
Q8_0
DeepSeek V4 Flash
291 GB302.3 GB file
32K 292 GB · 128K 293 GB
MiniMax M3
428 GB453.6 GB file
32K 430 GB · 128K 439 GB
16 more quants
IQ2_XXS
DeepSeek V4 Flash
80.3 GB90.9 GB file
32K 80.7 GB · 128K 81.8 GB
MiniMax M3
118 GB134.9 GB file
32K 121 GB · 128K 130 GB
IQ2_XS
DeepSeek V4 Flash
87.6 GB87.0 GB file
32K 88.0 GB · 128K 89.0 GB
MiniMax M3
129 GB129.5 GB file
32K 131 GB · 128K 140 GB
IQ2_S
DeepSeek V4 Flash
91.2 GB88.0 GB file
32K 91.6 GB · 128K 92.7 GB
MiniMax M3
134 GB132.1 GB file
32K 137 GB · 128K 146 GB
IQ2_M
DeepSeek V4 Flash
98.5 GB90.9 GB file
32K 98.9 GB · 128K 99.9 GB
MiniMax M3
145 GB135.1 GB file
32K 147 GB · 128K 157 GB
IQ3_XXS
DeepSeek V4 Flash
113 GB103.0 GB file
32K 113 GB · 128K 114 GB
MiniMax M3
166 GB160.3 GB file
32K 169 GB · 128K 178 GB
Q2_K_S
DeepSeek V4 Flash
117 GB98.6 GB file
32K 117 GB · 128K 118 GB
MiniMax M3
not listed
IQ3_XS
DeepSeek V4 Flash
not listed
MiniMax M3
177 GB196.4 GB file
32K 179 GB · 128K 189 GB
IQ3_S
DeepSeek V4 Flash
124 GB117.3 GB file
32K 124 GB · 128K 125 GB
MiniMax M3
182 GB175.6 GB file
32K 185 GB · 128K 194 GB
IQ3_M
DeepSeek V4 Flash
131 GB136.1 GB file
32K 132 GB · 128K 133 GB
MiniMax M3
193 GB205.5 GB file
32K 195 GB · 128K 205 GB
Q3_K_L
DeepSeek V4 Flash
not listed
MiniMax M3
219 GB205.0 GB file
32K 222 GB · 128K 231 GB
IQ4_XS
DeepSeek V4 Flash
157 GB137.9 GB file
32K 157 GB · 128K 158 GB
MiniMax M3
230 GB208.5 GB file
32K 233 GB · 128K 242 GB
IQ4_NL
DeepSeek V4 Flash
164 GB137.9 GB file
32K 164 GB · 128K 165 GB
MiniMax M3
241 GB212.7 GB file
32K 244 GB · 128K 253 GB
Q4_1
DeepSeek V4 Flash
not listed
MiniMax M3
241 GB268.9 GB file
32K 244 GB · 128K 253 GB
Q4_K_S
DeepSeek V4 Flash
not listed
MiniMax M3
241 GB248.6 GB file
32K 244 GB · 128K 253 GB
Q5_K_S
DeepSeek V4 Flash
not listed
MiniMax M3
294 GB299.7 GB file
32K 297 GB · 128K 306 GB
BF16
DeepSeek V4 Flash
582 GB~581.9 GB file
32K 583 GB · 128K 584 GB
MiniMax M3
855 GB852.9 GB file
32K 857 GB · 128K 866 GB
DeepSeek V4 Flash
Ollama · Q4_K_M
ollama run hf.co/teamblobfish/DeepSeek-V4-Flash-GGUF:Q4_K_MMiniMax M3
Ollama · Q4_K_M
ollama run hf.co/unsloth/MiniMax-M3-GGUF:Q4_K_M| Spec | DeepSeek V4 Flash | MiniMax M3 |
|---|---|---|
| Publisher | DeepSeek | MiniMax |
| Parameters | 290.9B | 427.0B |
| Active parameters | 14B per token | 20B per token |
| Architecture | DeepseekV4ForCausalLM | MiniMaxM3SparseForConditionalGeneration |
| Context | 1M tokens | 1M tokens |
| Modalities | Chat | Vision, Function-calling |
| License | MIT | Other |
| Released | Apr 2026 | Jun 2026 |
Frequently Asked Questions
Which is better, DeepSeek V4 Flash or MiniMax M3?
DeepSeek V4 Flash leads llmrun Score 115.3 to 112.6 (ranges overlap); neither fits 24 GB. Choose DeepSeek V4 Flash for higher llmrun Score and ahead on the benchmarks; choose MiniMax M3 for stronger at science and also does vision, function-calling.
Which needs less VRAM, DeepSeek V4 Flash or MiniMax M3?
At Q4_K_M, DeepSeek V4 Flash requires 174.9 GB while MiniMax M3 needs 256.7 GB, making DeepSeek V4 Flash the easier model to run on your local setup.
Can I run DeepSeek V4 Flash or MiniMax M3 on a 16 GB GPU?
DeepSeek V4 Flash does not fit (smallest quant needs 80.5 GB); MiniMax M3 does not fit (smallest quant needs 119.2 GB).
Which has a longer context window, DeepSeek V4 Flash or MiniMax M3?
DeepSeek V4 Flash supports 1,048,576 tokens and MiniMax M3 supports 1,048,576 tokens.