DeepSeek V4 Pro 0813 vs Muse Spark Open 816B A42B
DeepSeek V4 Pro 0813 and Muse Spark Open 816B A42B match in llmrun Score, with values of 129.2 versus 128.7; neither fits 24 GB.
Model A · DeepSeek
DeepSeek V4 Pro 08131650.5B · 102.3B active · 1M context · MIT · Aug 2026
Model B · Meta
Muse Spark Open 816B A42B822.0B · 42B active · Other · Sep 2026
Choose DeepSeek V4 Pro 0813 if you want…
Also does chat
Muse Spark Open 816B A42B doesn't list it
More permissive license
MIT vs Other
Choose Muse Spark Open 816B A42B if you want…
Lighter on memory
542.5 vs 990.6 GB at Q4_K_M
Also does vision
DeepSeek V4 Pro 0813 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 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
12 GB
RTX 3060 12GB · 12 GB tier
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
16 GB
RTX 5060 Ti 16GB · 16 GB tier
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
24 GBreference
RTX 4090 · 24 GB tier
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
32 GB
RTX 5090 · 32 GB tier
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
48 GB
RTX A6000 · 48 GB tier
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
Mac 32 GB
Mac Mini M4 · Macs
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
Mac 64 GB
Mac Studio M4 Max · Macs
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
Mac 128 GB
Mac Studio M4 Max · Macs
DeepSeek V4 Pro 0813
Doesn't fit
smallest quant needs 640.1 GB
Muse Spark Open 816B A42B
Doesn't fit
smallest quant needs 258.1 GB
Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).
Overall
DeepSeek V4 Pro 0813
Muse Spark Open 816B A42B
Coding
DeepSeek V4 Pro 0813
Muse Spark Open 816B A42B
no data
Math
DeepSeek V4 Pro 0813
Muse Spark Open 816B A42B
Science
DeepSeek V4 Pro 0813
Muse Spark Open 816B A42B
no data
Reasoning
DeepSeek V4 Pro 0813
Muse Spark Open 816B A42B
no data
DeepSeek V4 Pro 0813 · #5 of 94 open models · #5 of 93 that fit ≤1 TB
Muse Spark Open 816B A42B · #7 of 94 open models · #5 of 90 that fit ≤640 GB
Benchmarks
6 shared boards: DeepSeek V4 Pro 0813 leads 3, Muse Spark Open 816B A42B leads 3 · Benchmarks imported Oct 2026
DeepSeek V4 Pro 0813
91.7%#3/98 open
IndependentEpoch AI · Aug 2026
Muse Spark Open 816B A42B
88.6%#12/98 open
IndependentEpoch AI · Oct 2026
DeepSeek V4 Pro 0813
98.6%#1/86 open
IndependentEpoch AI · Aug 2026
Muse Spark Open 816B A42B
98.1%#3/86 open
IndependentEpoch AI · Oct 2026
DeepSeek V4 Pro 0813
64.6%#5/17 open
IndependentEpoch AI · Aug 2026
Muse Spark Open 816B A42B
69.8%#2/17 open
IndependentEpoch AI · Oct 2026
DeepSeek V4 Pro 0813
1582.0#11/52 open
ArenaLMArena · Oct 2026
Muse Spark Open 816B A42B
1511.3#13/52 open
ArenaLMArena · Oct 2026
DeepSeek V4 Pro 0813
52.9%#2/17 open
IndependentEpoch AI · Aug 2026
Muse Spark Open 816B A42B
56.6%#1/17 open
IndependentEpoch AI · Oct 2026
DeepSeek V4 Pro 0813
26.8%#6/13 open
IndependentEpoch AI · Aug 2026
Muse Spark Open 816B A42B
31.7%#2/13 open
IndependentEpoch AI · Oct 2026
17 more boards (other shared and single-model boards)
Only DeepSeek V4 Pro 0813 (17)
- LMArena Text1464.6
- LiveBench Coding77.2
- LiveBench Reasoning85.8
- LiveBench Math95.1
- ARC-AGI-261.3%
- SciCode51.0%
- ARC-AGI90.5%
- WeirdML66.2%
- FrontierCode28.5%
- Chess Puzzles47.0%
- CritPt18.0%
- DTBench93.9%
- LMCA45.5%
- APEX-Agents47.3%
- Mystery Game Puzzles43.0%
- ALE-Bench1403.2
- ProofBench50.0%
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 Pro 0813
702 GB587.3 GB file
32K 702 GB · 128K 704 GB
Muse Spark Open 816B A42B
384 GB304.1 GB file
32K 404 GB · 128K 469 GB
Q3_K_M
DeepSeek V4 Pro 0813
805 GB763.7 GB file
32K 805 GB · 128K 807 GB
Muse Spark Open 816B A42B
441 GB~400.7 GB file
32K 461 GB · 128K 525 GB
Q4_0
DeepSeek V4 Pro 0813
826 GB885.3 GB file
32K 826 GB · 128K 827 GB
Muse Spark Open 816B A42B
452 GB466.3 GB file
32K 472 GB · 128K 537 GB
Q4_K_M
DeepSeek V4 Pro 0813
991 GB~990.3 GB file
32K 991 GB · 128K 992 GB
Muse Spark Open 816B A42B
543 GB~493.2 GB file
32K 563 GB · 128K 627 GB
Q5_K_M
DeepSeek V4 Pro 0813
1176 GB~1176.0 GB file
32K 1177 GB · 128K 1178 GB
Muse Spark Open 816B A42B
644 GB~585.7 GB file
32K 664 GB · 128K 729 GB
Q6_K
DeepSeek V4 Pro 0813
1362 GB~1361.7 GB file
32K 1362 GB · 128K 1364 GB
Muse Spark Open 816B A42B
746 GB~678.2 GB file
32K 766 GB · 128K 831 GB
Q8_0
DeepSeek V4 Pro 0813
1651 GB~1650.5 GB file
32K 1651 GB · 128K 1653 GB
Muse Spark Open 816B A42B
904 GB871.6 GB file
32K 924 GB · 128K 989 GB
6 more quants
IQ2_XXS
DeepSeek V4 Pro 0813
not listed
Muse Spark Open 816B A42B
249 GB225.6 GB file
32K 269 GB · 128K 333 GB
IQ2_S
DeepSeek V4 Pro 0813
not listed
Muse Spark Open 816B A42B
283 GB275.5 GB file
32K 303 GB · 128K 367 GB
IQ3_XXS
DeepSeek V4 Pro 0813
640 GB619.5 GB file
32K 640 GB · 128K 642 GB
Muse Spark Open 816B A42B
350 GB340.4 GB file
32K 371 GB · 128K 435 GB
Q4_1
DeepSeek V4 Pro 0813
929 GB983.6 GB file
32K 929 GB · 128K 930 GB
Muse Spark Open 816B A42B
not listed
Q4_K_S
DeepSeek V4 Pro 0813
not listed
Muse Spark Open 816B A42B
509 GB470.1 GB file
32K 529 GB · 128K 593 GB
BF16
DeepSeek V4 Pro 0813
3301 GB~3301.0 GB file
32K 3302 GB · 128K 3303 GB
Muse Spark Open 816B A42B
1808 GB1640.3 GB file
32K 1829 GB · 128K 1893 GB
DeepSeek V4 Pro 0813
Ollama · Q4_K_M
ollama run hf.co/unsloth/DeepSeek-V4-Pro-0813-GGUFMuse Spark Open 816B A42B
Ollama · Q4_K_M
ollama run hf.co/unsloth/Muse-Spark-Open-816B-A42B-GGUF| Spec | DeepSeek V4 Pro 0813 | Muse Spark Open 816B A42B |
|---|---|---|
| Publisher | DeepSeek | Meta |
| Parameters | 1650.5B | 822.0B |
| Active parameters | 102B per token | 42B per token |
| Architecture | DeepseekV4ForCausalLM | — |
| Context | 1M tokens | — |
| Modalities | Chat | Vision |
| License | MIT | Other |
| Released | Aug 2026 | Sep 2026 |
Frequently Asked Questions
Which is better, DeepSeek V4 Pro 0813 or Muse Spark Open 816B A42B?
DeepSeek V4 Pro 0813 and Muse Spark Open 816B A42B match in llmrun Score, with values of 129.2 versus 128.7; neither fits 24 GB. Choose DeepSeek V4 Pro 0813 for also does chat and more permissive license; choose Muse Spark Open 816B A42B for lighter on memory and also does vision.
Which needs less VRAM, DeepSeek V4 Pro 0813 or Muse Spark Open 816B A42B?
At Q4_K_M, DeepSeek V4 Pro 0813 requires 990.6 GB while Muse Spark Open 816B A42B needs 542.5 GB, making Muse Spark Open 816B A42B the easier model to run on your local setup.
Can I run DeepSeek V4 Pro 0813 or Muse Spark Open 816B A42B on a 16 GB GPU?
DeepSeek V4 Pro 0813 does not fit (smallest quant needs 640.1 GB); Muse Spark Open 816B A42B does not fit (smallest quant needs 258.1 GB).