Which GPUs and Macs run MiniMax M3.1 Flash?
Key facts
Not on the cards on this page: MiniMax M3.1 Flash needs about 251 GB of memory, more than any of them has.
About 251.2 GB at UD-Q4_K_XL, our estimate. Checked . unsloth UD-Q4_K_XL GGUF.
None of the 43 graphics cards, PCs and Macs on our list holds the whole model at the standard quality in its graphics memory.
The smallest download, a file of 186.5 GB, does fit on one: Mac Studio M5 Ultra 256GB, about $11,565€12,500. It answers a little less well than the standard version.
EU shops · Shown to visitors in Europe.
US shops · Shown to visitors outside Europe.
If you buy through these links we may earn a commission. It doesn't change our ranking. How shop links work
As an Amazon Associate I earn from qualifying purchases.
Als Amazon-Partner verdiene ich an qualifizierten Verkäufen.
Als Amazon-Partner verdien ik aan in aanmerking komende aankopen.
We never pass off an estimate as a benchmark. Commissions never change which hardware we recommend. How we calculate
LIVE IN BROWSER · $0 · NO LOGIN
Test 10 open LLMs free in your browser before downloading 249 GB for MiniMax M3.1 Flash
Open full 8-model playground ↗Before downloading 249.4 GB of UD-Q4_K_XL weights for MiniMax M3.1 Flash (428B total, 23B active, MiniMax Community Licence with commercial-use conditions), you can test 10 open LLMs free in your browser at https://localllmrig.com/ with no login and 0 GB of local VRAM. Unsloth's 2-bit dynamic GGUF is 143.9 GB and runs in about 150 GB of RAM or unified memory by Unsloth's own figure. The prompts below run Recursive Flint and Gemma 4 31B. MiniMax M3.1 Flash is not one of the free models.
Test 10 open LLMs free in your browser →https://localllmrig.com/can-i-run/minimax-m3.1-flash
Click a starter prompt below or type your own question to watch Recursive Flint stream side by side against Gemma 4 31B right on this page:
Get a 1-line alert when a faster quant for MiniMax M3.1 Flash drops — or when GPUs that run it drop >10%
Get a 1-line alert when a new low-bit GGUF/MLX quant fits 128–256 GB rigs or when multi-GPU & Mac Studio prices drop >10% — weekly, zero spam.
Which graphics cards run it
Seven popular graphics cards and two PCs with two cards each, at the standard quality (Q4). Click one for the full answer: memory, speed and price.
Doesn't fit10
Not enough graphics memory: it needs about 251.2 GB.
- RTX 3060 12GBhas 12 GB
- RTX 4060 Ti 16GBhas 16 GB
- RTX 3090has 24 GB
- RTX 4090has 24 GB
- RTX 5080has 16 GB
- RX 9060 XT 16GBhas 16 GB
- RTX 5090has 32 GB
- RX 7900 XTXhas 24 GB
- Dual RTX 3090has 48 GB
- Dual RTX 4090has 48 GB
Which Macs and mini PCs run it
Macs and mini PCs from the picker, at the standard quality (Q4). A Mac shares its memory with the rest of the computer. Each name opens that machine's page.
Doesn't fit12
Not enough shared memory: it needs about 251.2 GB.
- Mac mini M4 16GBhas 16 GB
- Mac mini M6 32GBhas 32 GB
- Mac Studio M5 Max 36GBhas 36 GB
- Mac mini M4 Pro 64GBhas 64 GB
- Mac mini M5 Pro 64GBhas 64 GB
- GMKtec EVO-X2 (Ryzen AI Max+ 395) 64GBhas 64 GB
- Mac Studio M5 Ultra 96GBhas 96 GB
- GMKtec EVO-X2 (Ryzen AI Max+ 395) 128GBhas 128 GB
- GEEKOM A9 Mega (Ryzen AI Max+ 395) 128GBhas 128 GB
- Beelink GTR9 Pro (Ryzen AI Max+ 395) 128GBhas 128 GB
- Mac Studio M5 Max 128GBhas 128 GB
- Mac Studio M5 Ultra 256GBhas 256 GB
Cheapest hardware for each size
From all 43 graphics cards, PCs and Macs on our list. "Runs slowly" means part of the model runs from normal memory: cheaper, but much slower. A smaller download needs less memory and answers a little less well.
- Standard quality (UD-Q4_K_XL)No machine on our list runs this size.
- Standard quality (UD-Q4_K_XL), runs slowlyNo machine on our list runs this size.
- Smallest download (Text-3bit)Mac Studio M5 Ultra 256GB, about $11,565€12,500
- Smallest download (Text-3bit), runs slowlyNo machine on our list runs this size.
Prices excl. shipping unless marked. Euro prices incl. VAT where the seller charges it. Dollar prices are converted from our euro prices without the 21% EU VAT, before US sales tax.
Prices moved: see Building a local LLM rig in the RAMageddon.
About MiniMax M3.1 Flash
428B total, 23B active. Parameter count is an estimate. More parameters usually means a smarter model that needs more memory.
- Licence:
- minimax-community source
- Longest conversation:
- about 1573 pages of text (1,048,576 tokens)
- Input:
- text+image+video in, text out
- Released:
- source
Sources and method
MiniMax's open-weights multimodal mixture of experts: about 428B total, 23B active, context up to 1,048,576 tokens. Open weights under the MiniMax Community Licence. Non-commercial use is free; commercial use needs a credit and a notice to MiniMax, or written permission above US$20 million yearly revenue. This is not an Apache-2.0 licence. On 26 Oct 2026 Unsloth published dynamic GGUFs (https://huggingface.co/unsloth/MiniMax-M3.1-Flash-GGUF, guide https://unsloth.ai/docs/models/minimax-m3.1-flash). UD-Q2_K_XL is 143.9 GB. Unsloth says that 2-bit file runs in about 150 GB of RAM or unified memory, with 91.5% top-1 token agreement, for a 16K context and one slot (Unsloth, with day-zero weights from MiniMax). A 192 GB, 256 GB, or 512 GB Mac Studio, a dual DGX Spark, or an EPYC server with 256 GB of DDR4 or DDR5 is the class that figure points at. UD-Q4_K_XL is 249.4 GB. Unsloth's planning figure for that 4-bit file is 256 GB or more of unified memory, which also covers an eight-node DGX Spark. Official NVFP4 and MXFP8 weights are 252.42 GB and are not a GGUF. Text-only MLX builds remain listed, the smallest of those at 186.49 GB. Our conversation-memory estimate keeps a full cache for every layer; the sparse indexer cache is not modelled.
Ran MiniMax M3.1 Flash on your own hardware? Send us your llama-bench result.
Memory at each quality level
Each download of MiniMax M3.1 Flash is a different size. The total adds the file, the memory a conversation of about 12 pages of text (8,192 tokens) takes, and 0.8 GB of working space. A measured file size is labelled; the rest is our estimate. The last column is the cheapest machine that holds the total.
Swipe the table sideways to see every column, up to the cheapest machine.
| Quality (file) | File size | Conversation | Working space | Total | Cheapest machine that holds it |
|---|---|---|---|---|---|
| UD-Q8_K_XL | 259.10 GBUnsloth: lossless conversion of the NVFP4 and BF16 weights, 26 Oct 2026 | 1.0 GB | 0.8 GB | 260.9 GB | No machine on our list runs this size. |
| UD-Q4_K_XL | 249.40 GBmeasured GGUF | 1.0 GB | 0.8 GB | 251.2 GB | No machine on our list runs this size. |
| oQ4e-mtpMLX: Apple Silicon only | 248.77 GBmeasured MLX | 1.0 GB | 0.8 GB | 250.6 GB | No machine on our list runs this size. |
| Text-4bitMLX: Apple Silicon only | 248.37 GBmeasured MLX | 1.0 GB | 0.8 GB | 250.2 GB | No machine on our list runs this size. |
| oQ3.5e-mtpMLX: Apple Silicon only | 213.37 GBmeasured MLX | 1.0 GB | 0.8 GB | 215.2 GB | No machine on our list runs this size. |
| oQ3e-mtpMLX: Apple Silicon only | 196.15 GBmeasured MLX | 1.0 GB | 0.8 GB | 198.0 GB | No machine on our list runs this size. |
| Text-3bitMLX: Apple Silicon only | 186.49 GBmeasured MLX | 1.0 GB | 0.8 GB | 188.3 GB | Mac Studio M5 Ultra 256GB, $11,565€12,500, price unverified |
| UD-Q2_K_XL | 143.90 GBUnsloth: 91.5% top-1 token agreement, about 150 GB RAM or unified memory, 26 Oct 2026 | 1.0 GB | 0.8 GB | 145.7 GB | Mac Studio M5 Ultra 256GB, $11,565€12,500, price unverified |
speed: not estimated for MoE yet.
Speeds and memory are our estimate from the hardware specs, not a measured test. Prices are dated and may change. How we calculate
Questions
- Can I run MiniMax M3.1 Flash on 24GB?
- Not in full. MiniMax M3.1 Flash (428B total, 23B active, parameter count is an estimate) needs about 251.2 GB at the standard quality (our estimate, 249.4 GB for the weights + 1.01 GB conversation memory + 0.80 GB working space), and a 24 GB card leaves about 23.4 GB usable. With the measured oQ4e-mtp MLX build the total is 250.6 GB (248.77 GB file).
- Can I run MiniMax M3.1 Flash on a 128GB Mac?
- No, by our estimate. MiniMax M3.1 Flash needs about 251.2 GB at the standard quality and 188.3 GB even at Text-3bit (186.49 GB measured MLX), while Mac Studio M4 Max 128GB and Mac Studio M5 Max 128GB leave about 96.0 GB usable of 128 GB.
- Which Mac or server can run MiniMax M3.1 Flash locally?
- The smallest download on our list, Text-3bit (186.49 GB measured MLX, 188.3 GB total estimate) at 8,192 tokens, fits on Mac Studio M5 Ultra 256GB (192.0 GB usable of 256 GB, €12,500 / $11,565, price unverified). The 4-bit MLX builds (oQ4e-mtp (248.77 GB measured MLX, 250.6 GB total estimate) and Text-4bit (248.37 GB measured MLX, 250.2 GB total estimate)) are larger than the 192.0 GB usable on the largest Mac on our list (Mac Studio M5 Ultra 256GB). The model notes record official NVFP4 and MXFP8 weights of 252.42 GB, which are not a GGUF. On 26 Oct 2026 Unsloth published dynamic GGUFs. UD-Q2_K_XL (143.9 GB measured GGUF, 145.7 GB total estimate) is the 2-bit file. Unsloth says it runs in about 150 GB of RAM or unified memory, with 91.5% top-1 token agreement, for a 16K context and one slot (Unsloth, with day-zero weights from MiniMax; not a measurement on this site). A 192 GB, 256 GB, or 512 GB Mac Studio, a dual DGX Spark, or an EPYC server with 256 GB of DDR4 or DDR5 is the class that figure points at. UD-Q4_K_XL (249.4 GB measured GGUF, 251.2 GB total estimate) is the 4-bit dynamic file. Unsloth's planning figure for it is 256 GB or more of unified memory, which also covers an eight-node DGX Spark.
- Can I test MiniMax M3.1 Flash or comparable open models online in my browser first?
- MiniMax M3.1 Flash is not one of the 10 free open models (26B to 280B). You can test those free models, including Gemma 4 31B (https://localllmrig.com/can-i-run/gemma-4-31b), Gemma 4 26B A4B (https://localllmrig.com/can-i-run/gemma-4-26b-a4b), NVIDIA Nemotron 3.5 Lightning (https://localllmrig.com/can-i-run/nemotron-3.5-lightning), Laguna S 2.1 (https://localllmrig.com/can-i-run/laguna-s-2.1), NVIDIA Nemotron 3.5 Super VL (https://localllmrig.com/can-i-run/nemotron-3.5-super-vl), Recursive Flint (https://localllmrig.com/can-i-run/flint), for $0 with no account at https://localllmrig.com/ or in the inline test box on https://localllmrig.com/can-i-run/minimax-m3.1-flash. MiniMax M3.1 Flash (428B, 23B active) is in the chat catalog and is not one of the free models, so trying it needs a pass at https://localllmrig.com/. MiniMax M3 (428B, 23B active) is in the chat catalog and is not one of the free models, so trying it needs a pass at https://localllmrig.com/.
- How fast is MiniMax M3.1 Flash on an RTX 5090?
- We don't give a speed for this model yet. This model only uses part of itself for each word (23B active of 428B total, a mixture of experts), and our speed estimate does not handle that yet.
- What is the license and maximum context length of MiniMax M3.1 Flash?
- MiniMax M3.1 Flash is an open-weights model (text+image+video in, text out) and a mixture of experts, released 17 Oct 2026, with a maximum context length of 1,048,576 tokens (about 1,573 pages of text). Licence: minimax-community. Open weights. Non-commercial use is free. Commercial use needs a credit and a notice to MiniMax, or written permission above US$20M yearly revenue.