DiffusionGemma 26B A4B needs ~18.1 GB VRAM. RTX 4000 Ada 20GB has 20.0 GB. With Q2_K quantization, expect ~45 tok/s.
Operating mode
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
3.8 GB over capacity — needs offload or smaller quantization
Fit status
Too heavy
Decode
17.5 tok/s
TTFT
11087 ms
Safe context
4K
Memory
23.8 GB / 20.0 GB
Offload
20%
This setup is broadly balanced for this model.
No major red flags
This recommendation has enough memory headroom and acceptable estimated speed for the selected workload.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | F | Too heavy | 20.7 tok/s | 5110 ms | 4K |
| Coding | F | Too heavy | 17.5 tok/s | 11087 ms | 4K |
| Agentic Coding | F | Too heavy | 12.9 tok/s | 21797 ms | 4K |
| Reasoning | F | Too heavy | 17.5 tok/s | 13103 ms | 4K |
| RAG | F | Too heavy | 12.9 tok/s | 27246 ms | 4K |
Inference speed
Estimated decode speed (tokens/sec) for DiffusionGemma 26B A4B at Q4_K_M across popular GPUs and Apple Silicon, using the fastest local runtime per device. Fastest is RTX 5090 32GB at ~144 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.
| GPU / Mac | Memory | Quant | Speed (tok/s) | Fits? |
|---|---|---|---|---|
 RTX 5090 32GB | 32 GB | Q4_K_M | 143.6 | Fits |
| RTX 4090 24GB | 24 GB | Q4_K_M | 67.5 | Too big |
Mac Studio M3 Ultra 256GB | 256 GB | Q4_K_M | 66.6 | Fits |
RX 7900 XTX 24GB | 24 GB | Q4_K_M | 60.9 | Too big |
| RTX 3090 24GB | 24 GB | Q4_K_M | 57.7 | Too big |
Mac Studio M2 Ultra 128GB | 128 GB | Q4_K_M | 55.5 | Fits |
Mac Studio M1 Ultra 128GB | 128 GB | Q4_K_M | 52.6 | Fits |
MacBook Pro M4 Max 128GB | 128 GB | Q4_K_M | 41.1 | Fits |
MacBook Pro M4 Max 64GB | 64 GB | Q4_K_M | 41.1 | Fits |
MacBook Pro M3 Max 64GB | 64 GB | Q4_K_M | 28.7 | Fits |
MacBook Pro M1 Max 64GB | 64 GB | Q4_K_M | 26.3 | Fits |
MacBook Pro M4 Pro 48GB | 48 GB | Q4_K_M | 25.1 | Fits |
| RTX 4080 Super 16GB | 16 GB | Q4_K_M | 24.6 | Too big |
| RTX 4070 12GB | 12 GB | Q4_K_M | 8.6 | Too big |
| RTX 3060 12GB | 12 GB | Q4_K_M | 5.4 | Too big |
| RTX 4060 8GB | 8 GB | Q4_K_M | 3.6 | Too big |
Estimates for single-stream decoding at Q4_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.
How DiffusionGemma 26B A4B (25.799999237060547B params) fits at each quantization level on RTX 4000 Ada 20GB (20.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 10.1 GB | Low | A80 |
Q3_K_S | 3 | 12.6 GB | Low | A79 |
NVFP4Best for your GPU | 4 | 14.4 GB | Medium | A79 |
Q4_K_M | 4 | 15.7 GB | Medium | F0 |
Q5_K_M | 5 | 18.6 GB | High | F0 |
Q6_K | 6 | 21.2 GB | High | F0 |
Q8_0 | 8 | 27.6 GB | Very High | F0 |
F16 | 16 | 52.9 GB | Maximum | F0 |
Copy-paste commands to run DiffusionGemma 26B A4B on your machine.
Run
docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
--hf-repo "google/diffusiongemma-26B-A4B-it" \
--hf-file "diffusiongemma-26B-A4B-it-Q4_K_M.gguf" \
-c 4096 -ngl 99Opções de upgrade
RTX 5090 32GBOpção econômicaMakes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$1,999 MSRP
RTX PRO 4500 Blackwell 32GBMelhor custo-benefícioMakes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$2,499 MSRP
RTX 5000 Ada 32GBUpgrade NVIDIAMakes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$4,000 MSRP
Yes, RTX 4000 Ada 20GB can run DiffusionGemma 26B A4B at Q2_K quantization (Tight fit). The recommended Q4_K_M requires 23.8 GB which exceeds available memory, but at Q2_K it needs only 18.1 GB. Expected decode speed: 44.6 tok/s.
DiffusionGemma 26B A4B (25.799999237060547B parameters) requires approximately 23.8 GB at Q4_K_M quantization. On RTX 4000 Ada 20GB, it fits at Q2_K using 18.1 GB.
The recommended quantization is Q4_K_M, but on RTX 4000 Ada 20GB the best fitting quantization is Q2_K, which uses 18.1 GB.
On RTX 4000 Ada 20GB, DiffusionGemma 26B A4B achieves approximately 44.6 tokens per second decode speed with a time-to-first-token of 4337ms using Q2_K quantization.
For coding workloads, DiffusionGemma 26B A4B on RTX 4000 Ada 20GB receives a F grade with 17.5 tok/s and 4K context.
On RTX 4000 Ada 20GB, DiffusionGemma 26B A4B can safely use up to 24K tokens of context at Q2_K quantization. The model's official context limit is 262K, but available memory constrains the safe maximum.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/diffusiongemma-26b-a4b-on-rtx-4000-ada-20gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: