SAVRN
Search Contact SAVRN

SAVRN Model Hub · Comparisons

DeepSeek-V4-Flash-0731 vs GLM-5.2-NVFP4

DeepSeek-V4-Flash-0731 has 304.2B parameters and GLM-5.2-NVFP4 has 381B parameters; both are released under MIT License; at 16-bit, DeepSeek-V4-Flash-0731 needs about 730 GB (3x MI325X from $6.00 an hour) and GLM-5.2-NVFP4 about 914.4 GB (4x MI325X from $8.00 an hour).

Published metadata for 2 models, each read from its own repository.
Field DeepSeek-V4-Flash-0731
deepseek-ai/DeepSeek-V4-Flash-0731
GLM-5.2-NVFP4
nvidia/GLM-5.2-NVFP4
Publisher DeepSeek NVIDIA
Task Text generation Text generation
Modality Text Text
Parameters, as reported 304.2B parameters 381B parameters
Architecture DeepseekV4ForCausalLM GlmMoeDsaForCausalLM
Library transformers Model Optimizer
Context length 1,048,576 tokens 1,048,576 tokens
Repository size 166.9 GB 464.9 GB
Artifact formats safetensors safetensors
License mit mit
Access Open weights, no gate Open weights, no gate
Memory at 16-bit (weights and margin) 730 GB 914.4 GB
Cheapest GPUs at 16-bit, per hour 3x MI325X, $6.00 4x MI325X, $8.00
Memory at 4-bit (weights and margin) 182.5 GB 228.6 GB
Cheapest GPUs at 4-bit, per hour 1x MI300X, $1.85 1x MI325X, $2.00
Revision viewed 7872f01b1d1f 53e0691e2189
Downloads reported by the hub 4.3M 859.8k
Last observed 2026-09-18 2026-09-18

An evaluation row appears only where at least two of these models report the same benchmark with the same stated configuration, metric, unit and setup. Different evaluators stay named in each cell. Values are shown as reported: no unit conversion, no ranking.

Other Reported Results

These results are listed for each model on its own, because the conditions needed to compare them are not stated or do not match. Two results that leave a condition blank are not assumed to share it.

DeepSeek-V4-Flash-0731

BenchmarkConditionsResultReported byRevisionDate
datacurve/deep-swe Task deep_sweMetric deep_sweComparison conditions not established 54.4 Model Card
Reported by a third party
Evaluated revision not stated 2026-08-03
harborframework/terminal-bench-2.1 Task terminalbench_2_1Metric terminalbench_2_1Setup DeepSeek Harness (minimal mode), max reasoning effort, temperature=1.0, top_p=0.95.Comparison conditions not established 82.7 Model Card
Reported by a third party
Evaluated revision not stated 2026-08-01
hkust-nlp/Toolathlon Task toolathlon_verifiedMetric toolathlon_verifiedComparison conditions not established 70.3 deepseek-ai/DeepSeek-V4-Flash-0731 model card
Reported by a third party
Evaluated revision not stated 2026-08-01

SAVRN's Notes on DeepSeek-V4-Flash-0731

Only 6 of the 256 routed experts fire on any given token, but all 304.2B parameters have to be resident, and that sets the bill for DeepSeek-V4-Flash-0731. At 16-bit the working set is 730 GB: three MI325X cards with 256 GB each at $6.00 an hour. At 8-bit, 365 GB fits on two MI300X at $3.70; at 4-bit, 182.5 GB fits on one MI300X at $1.85. It generates text over a 1,048,576-token context.

SAVRN Index host prices run from $0.06 in and $0.18 out per million tokens at DeepInfra to $0.44 and $1.32 at Novita, more than a seven-fold spread, so price a run yourself. MIT keeps that simple: commercial use, modification and redistribution, provided the copyright and permission notice travels with the files. Decide how much of the million-token window you will run, since the memory figures are quoted on the weights, and read arXiv:2606.19348, the paper behind it.

SAVRN's Notes on GLM-5.2-NVFP4

We read this build as the way to get GLM-5.2 onto one card. NVIDIA's NVFP4 quantization of ZAI's 381B-parameter mixture-of-experts model, 256 routed experts, built for reasoning and coding, needs 228.6 GB at 4-bit and fits one MI325X with 256 GB at $2.00 an hour on demand. At 16-bit the same model needs 914.4 GB and four of those cards at $8.00 an hour. Plan disk separately: the 56 safetensors files total 464.9 GB.

The MIT terms match the base model and allow commercial use, modification and redistribution with the notices kept. Before you buy hardware for it, check how much of the 1,048,576-token window you can serve on one 256 GB card with 190.5 GB of weights already loaded, confirm your inference stack handles the GlmMoeDsaForCausalLM architecture with its sparse attention, and compare against the unquantized zai-org/GLM-5.2 if 4-bit is not your target.

Questions

Which is larger, DeepSeek-V4-Flash-0731 or GLM-5.2-NVFP4?

GLM-5.2-NVFP4 (381B parameters) is larger than DeepSeek-V4-Flash-0731 (304.2B parameters), by the parameter counts their publishers report.

Which is cheaper to run, DeepSeek-V4-Flash-0731 or GLM-5.2-NVFP4?

At 4-bit, DeepSeek-V4-Flash-0731 fits on 1x MI300X from $1.85 an hour and GLM-5.2-NVFP4 on 1x MI325X from $2.00 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use DeepSeek-V4-Flash-0731 commercially?

Yes. DeepSeek-V4-Flash-0731 is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.

Can I use GLM-5.2-NVFP4 commercially?

Yes. GLM-5.2-NVFP4 is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.

Related Comparisons