How much VRAM does Jamba v0.1 52B (12B active) need?
Jamba v0.1 52B (12B active) needs about 31.3 GB of VRAM at Q4_K_M, or 54.4 GB at Q8_0 and 101.7 GB at FP16. No consumer card holds it at Q4_K_M. The smallest fit in our set is the A100 40GB (40 GB). Weights plus runtime overhead. Budget separately for KV cache, which at 256K is a material share.
Weights-plus-runtime footprint across the six most common quantizations. KV cache is context-dependent and comes on top. At 256K it is gigabytes on its own, before the weights; size it precisely in the GPU Memory Calculator.
This is a mixture-of-experts model: all 51.6B parameters must sit in memory, but only about 12B are active per token — so it loads like a 51.6B model and runs closer to a 12B one.
This is the base checkpoint: pretrained on raw text and not tuned to follow instructions, so it continues text rather than answering prompts. That makes it the right starting point if you plan to fine-tune your own model. For chat or assistant work, use Jamba 1.5 Mini 52B Instruct — same parameter count, so the memory figures below are identical.
Can your GPU run Jamba v0.1 52B (12B active)?
Pick your card — the answer below is computed by the same engine that produces every figure on this page.
A100 40GB runs Jamba v0.1 52B (12B active) at Q4_K_M with 6.7 GB to spare; room left for context and everyday runtime pressure; as a mixture-of-experts model all 51.6B parameters must be resident even though only 12B are active per token.
Weights plus runtime overhead. KV cache is not included — it depends on context length and this model’s published architecture.
| Weights (Q4_K_M) | 29.1 GB |
| Runtime overhead | 2.2 GB |
| Estimated total | 31.3 GB |
| Usable VRAM (95% of 40 GB) | 38.0 GB |
Across every GPU we track
- RTX 3060 12GBNot enough-19.9 GB
- RTX 4060 Ti 16GBNot enough-16.1 GB
- RTX 4070 Ti SUPERNot enough-16.1 GB
- RTX 3090Not enough-8.5 GB
- RTX 4090Not enough-8.5 GB
- RTX 5090Not enough-0.9 GB
- A100 40GBRecommended+6.7 GB
- RTX A6000Recommended+14.3 GB
- L40SRecommended+14.3 GB
- A100 80GBExcellent+44.7 GB
- H100 80GBExcellent+44.7 GB
- H200 141GBExcellent+102.6 GB
Estimates, not guarantees: real usage moves with runtime, driver, batch size and context. Size a specific context window in the GPU Memory Calculator.
How was this number calculated? Every figure here comes from Bitpute’s documented calculation methodology — parameters, precision, quantization, runtime overhead and usable VRAM, each formula written out in full.
Memory by quantization
| Quant | Weights | + runtime overhead |
|---|---|---|
| FP16 | 96.1 GB | 101.7 GB |
| Q8_0 | 51.1 GB | 54.4 GB |
| Q6_K | 39.4 GB | 42.1 GB |
| Q5_K_M | 34.2 GB | 36.6 GB |
| Q4_K_M | 29.1 GB | 31.3 GB |
| Q4_0 | 27.0 GB | 29.1 GB |
Overhead = 0.75 GB + 5% of weights (CUDA context, buffers). Add KV cache on top. Weights dominate at this size, but a 256K window still costs gigabytes.
Which quantization should you actually run?
For Jamba v0.1 52B (12B active), the single-GPU entry point is Q4_K_M (31.3 GB) on RTX A6000, with plenty of headroom for long context. Higher quality (Q6_K 42.1 GB, FP16 101.7 GB) needs a 48 GB+ card or multiple GPUs — for occasional use, renting is usually cheaper than buying. As a mixture-of-experts model it must hold all 51.6B in VRAM but activates only about 12B per token, so it runs much faster than its footprint suggests — memory is the limit here, not speed.
Single-GPU fit at Q4_K_M (31.3 GB + KV)
RTX 3060 12GBRTX 4060 Ti 16GBRTX 4070 Ti SUPERRTX 3090RTX 4090RTX 5090A100 40GBRTX A6000L40SA100 80GBH100 80GBH200 141GB
Where the memory goes
Weights and overhead are exact for Q4_K_M. The KV bar is a generic 32-layer transformer at 8K; at this scale the weights dominate, but long context still adds gigabytes — size it in the calculator is for.
Deciding how much headroom to buy at this tier? A100 40GB vs RTX A6000 compares them on bandwidth, power and which models each one holds.
Common questions
How much VRAM does Jamba v0.1 52B (12B active) need?
Jamba v0.1 52B (12B active) needs about 31.3 GB of VRAM at Q4_K_M, 54.4 GB at Q8_0, or 101.7 GB at FP16. That is weights plus about 0.75 GB of runtime overhead and 5% of weight size; KV cache is additional and depends on context length.
What GPU can run Jamba v0.1 52B (12B active)?
At Q4_K_M the smallest card in our set that fits is the A100 40GB (40 GB usable at 95%). At FP16 you need the H200 141GB (141 GB) or larger.
Can Jamba v0.1 52B (12B active) run on 24 GB?
Not at Q4_K_M: it needs about 31.3 GB, more than the ~22.8 GB usable on a 24 GB card. Use a larger card, multiple GPUs, or offload.
Same memory footprint as 1 other checkpoint
Jamba v0.1 52B (12B active) has an identical parameter count to this, so every figure on this page applies to it unchanged. What varies is different checkpoint roles (instruction-tuned) — not the size. Choose on capability and licence; the memory budget is identical.
More in the Jamba family
Jamba 1.5 Mini 52B InstructJamba 1.5 Large 398B (94B active) Instruct