DeepSeek V4 Flash 0731

At the default Q4_K_XL, 0 of 139 phones have enough estimated usable memory; 0 also reach 3+ tokens/s. On Mac, 11 of 54 exact configurations pass using that quant or a smaller tracked fallback.

Params
284B (13B active)
Family
deepseek
Released
2026-07
Tags
chat · coding · reasoning

128GB MAC STUDIO RESULT

M4 Max 128GB fits the IQ3_XXS build

The default Q4_K_XL weights do not fit this configuration. The checker falls back to the tracked 104.2GB IQ3_XXS build: about 110.4GB of working memory and an estimated 825.2 tokens/s. This is a formula range for the listed GGUF, not a measured claim; custom DS4 and oMLX formats are not directly comparable.

Open the exact Mac Studio report →

Downloads & memory by quant

The download is only the weights — add KV cache and runtime to get what your phone actually needs. “Memory-fit” only means it can load; “usable” also requires an estimated 3+ tokens/s. Each row uses that row's quant. ★ = recommended.

QuantDownloadMemory neededPhone reality
IQ1_S82.5 GB~87.4 GB0 memory-fit
IQ1_M86.9 GB~92 GB0 memory-fit
IQ2_M90.9 GB~96.2 GB0 memory-fit
IQ2_XXS90.9 GB~96.2 GB0 memory-fit
Q2_K_XL96.8 GB~102.4 GB0 memory-fit
IQ3_XXS104.2 GB~110.2 GB0 memory-fit
IQ3_S116.1 GB~122.7 GB0 memory-fit
Q3_K_M128.1 GB~135.3 GB0 memory-fit
Q3_K_XL128.2 GB~135.4 GB0 memory-fit
IQ4_NL136.7 GB~144.3 GB0 memory-fit
IQ4_XS136.7 GB~144.3 GB0 memory-fit
Q4_K_XL155.1 GB~163.6 GB0 memory-fit
Q8_K_XL161.9 GB~170.8 GB0 memory-fit

DeepSeek V4 Flash 0731 at Q4_K_XL on 139 phones

This table is fixed to the recommended Q4_K_XL, on each phone's largest RAM option. Counts for smaller quants above will differ. All phones are shown, sorted by fit and estimated experience; tap one for the full report.

PhoneChip · RAMNeedsSpeedVerdict

~ = bandwidth-based estimate · ✓ = measured on real hardware

DeepSeek V4 Flash 0731 on 54 Mac configurations

Each Mac uses the recommended Q4_K_XL at 4K context when it fits, or the largest smaller tracked quant that fits. Each row links to the full report with a macOS-specific setup guide.

MacUnified memoryQuantWorking budgetNeedsEstimated speedVerdict

FAQ

How much RAM do I need to run DeepSeek V4 Flash 0731 on a phone?

At Q4_K_XL, DeepSeek V4 Flash 0731 needs ~163.6GB of usable memory (weights + KV cache + runtime). In practice that means no current Android phone or no current iPhone — iOS lets a single app use less of its RAM than Android does.

How fast does DeepSeek V4 Flash 0731 run on a flagship phone?

It doesn't fit on any of the 139 phones we track at Q4_K_XL — it needs ~163.6GB of usable memory.

Can an iPhone run DeepSeek V4 Flash 0731?

Not really: the best iPhone we track (iPhone 16 Pro Max) has ~5.2GB usable, but DeepSeek V4 Flash 0731 needs 163.6GB at Q4_K_XL.

What is the best quantization of DeepSeek V4 Flash 0731 for mobile?

Q4_K_XL (155.1GB download) is the size/quality sweet spot of the 13 quants available. Total memory needed is ~163.6GB once the KV cache and runtime are counted — the download size alone understates it.

How many phones can run DeepSeek V4 Flash 0731?

0 of the 139 phones we track have enough estimated usable memory at Q4_K_XL; 0 also reach our usable-speed threshold of 3 tokens/s, and 0 run smoothly. Memory-fit alone does not mean a good experience.

Which Macs can run DeepSeek V4 Flash 0731?

11 of the 54 exact Mac configurations we track have enough estimated working memory. Each Mac uses the recommended Q4_K_XL when it fits, or the largest smaller tracked quant that fits, and separates installed unified memory, the macOS reserve, and estimated speed.

What is DeepSeek V4 Flash 0731 good for on a phone?

It's tagged for chat, coding, reasoning. At 284B parameters (13B active — it's a MoE, so it decodes faster than its size suggests), it prioritizes answer quality over speed — expect slower decoding.

Full requirements guide
How can I run DeepSeek Harness locally? Tested setup, real screenshots, and local-model configuration

Guides for DeepSeek

DeepSeek Harness Has 4 Modes. Which One Should You Use?
DeepSeek Harness vs Claude Code & Codex: Which Fits?

Similar models

DeepSeek R1 Distill 1.5BDeepSeek R1 Distill 7BInkling SmallHunyuan 3 (Hy3)Ornith 1.0 397B