DiffusionGemma 26B A4B needs ~25.0 GB VRAM. AMD Instinct MI100 32GB has 32.0 GB. With Q4_K_M quantization, expect ~95 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
Fit status
Runs well
Decode
95.4 tok/s
TTFT
2028 ms
Safe context
47K
Memory
25.0 GB / 32.0 GB
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 | S | Runs well | 95.4 tok/s | 1106 ms | 47K |
| Coding | S | Runs well | 95.4 tok/s | 2028 ms | 47K |
| Agentic Coding | A | Tight fit | 95.4 tok/s | 2950 ms | 47K |
| Reasoning | S | Runs well | 95.4 tok/s | 2397 ms | 47K |
| RAG | A | Tight fit | 95.4 tok/s | 3688 ms | 47K |
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 AMD Instinct MI100 32GB (32.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 10.1 GB | Low | A75 |
Q3_K_S | 3 | 12.6 GB | Low | A76 |
NVFP4 | 4 | 14.4 GB | Medium | A77 |
Q4_K_M | 4 | 15.7 GB | Medium | A78 |
Q5_K_M | 5 | 18.6 GB | High | A78 |
Q6_KBest for your GPU | 6 | 21.2 GB | High | A78 |
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 99Your hardware
| Model | Params | Grade | Decode | Capabilities |
|---|---|---|---|---|
 Qwen3-Coder 30B A3B Instruct | 30.5B | S | 91.7 tok/s | |
| Qwen 3.5 27B | 27B | S | 41.9 tok/s | |
| Qwen 3.6 27B | 27B | S | 42 tok/s | |
| Qwen3-VL 30B A3B Instruct | 30B | S | 94.9 tok/s | |
| Qwen 3.5 35B A3B | 35B | S | 83.8 tok/s |
Yes, AMD Instinct MI100 32GB can run DiffusionGemma 26B A4B with a S grade (Runs well). Expected decode speed: 95.4 tok/s.
DiffusionGemma 26B A4B (25.799999237060547B parameters) requires approximately 25.0 GB of memory with Q4_K_M quantization.
The recommended quantization for DiffusionGemma 26B A4B is Q4_K_M, which balances quality and memory efficiency.
On AMD Instinct MI100 32GB, DiffusionGemma 26B A4B achieves approximately 95.4 tokens per second decode speed with a time-to-first-token of 2028ms using Q4_K_M quantization.
For coding workloads, DiffusionGemma 26B A4B on AMD Instinct MI100 32GB receives a S grade with 95.4 tok/s and 47K context.
On AMD Instinct MI100 32GB, DiffusionGemma 26B A4B can safely use up to 47K tokens of context. 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-instinct-mi100-32gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: