Can the RTX 3090 run WeMM-Video-35BA3B?
WeMM-Video-35BA3B·36.2B (5B active)·GeForce RTX 3090·24 GB VRAM
Yes. The RTX 3090 runs WeMM-Video-35BA3B fully in VRAM at Q4_K_M (about 23 GB of 24 GB), at roughly 171 tokens/s.
- best quant
- Q4_K_M
- needs
- 23 GB
- usable
- 24 GB
- est. speed
- ~171 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the WeMM-Video-35BA3B page →
Every quantisation of WeMM-Video-35BA3B on the RTX 3090
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 13 GB | fits | ~315 tok/s |
| Q3_K_M | 18 GB | fits | ~212 tok/s |
| Q4_K_M pick | 23 GB | fits | ~171 tok/s |
| Q5_K_M | 26 GB | offload | ~64 tok/s |
| Q6_K | 30 GB | offload | ~31 tok/s |
| Q8_0 | 39 GB | offload | ~15 tok/s |
| FP16 / BF16 | 72 GB | too big | — |
Speed is a bandwidth-based estimate for generating text (936 GB/s); prompt processing and long contexts are slower.
Measured speed on the RTX 3090
No community reports for this pair yet.
Report your own speed →Vision models that fit the RTX 3090 comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 36BA3B | Q4_K_M | ≈22 GBFits in memory | ~283 tok/s | |
| 36BA3B | Q4_K_M | ≈23 GBFits in memory | ~283 tok/s | |
| 27.8B | Q5_K_M | ≈22 GBFits in memory | ~26 tok/s | |
| 35.2BA3B | Q4_K_M | ≈22 GBFits in memory | ~283 tok/s | |
| 29.8B | Q5_K_M | ≈22 GBFits in memory | ~24 tok/s | |
| 27.8B | Q5_K_M | ≈22 GBFits in memory | ~26 tok/s |
Other GPUs that run WeMM-Video-35BA3B
Frequently asked questions
- Can the RTX 3090 run WeMM-Video-35BA3B?
- Yes. The RTX 3090 runs WeMM-Video-35BA3B fully in VRAM at Q4_K_M (about 23 GB of 24 GB), at roughly 171 tokens/s.
- How fast does WeMM-Video-35BA3B run on the RTX 3090?
- An estimate from memory bandwidth gives about 171 tokens/s at Q4_K_M.
- Which quantisation of WeMM-Video-35BA3B should I use on the RTX 3090?
- Q4_K_M: the largest one that fits. Expected quality: small loss compared with the full-precision weights.
Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.