What GPU do I need to run RedHatAI/Qwen3-8B-FP8-dynamic?
8.2B parameters, published in F8_E4M3. View on Hugging Face
Set up Qwen3-8B-FP8-dynamicQwen3-8B-FP8-dynamic is published by RedHatAI on Hugging Face, with 14,327 downloads and 14 likes to date. It's a Qwen3ForCausalLM model built for text-generation, published natively in F8_E4M3.
VRAM required & cheapest live GPU fit
Required VRAM = weight size at each precision times a flat 1.2 overhead for activations and fragmentation. The KV-cache grows with context and is not in that factor; it is listed per model below where the architecture is published. Full formula and assumptions: methodology.
| Precision | Weight size | Required VRAM | Cheapest live fit | GPUs needed | Est. $/hr (full fit) |
|---|---|---|---|---|---|
| FP8 (native) | 7.6 GB | 9.2 GB | RTX 4070 Super | 1 | $0.121/hr |
| INT4 (quantized) | 3.8 GB | 4.6 GB | RTX 4070 Super | 1 | $0.121/hr |
A GPU is only matched to a row if its hardware supports that precision, and the primary recommendation is always a single-GPU fit when one exists.
INT4 requires a quantized checkpoint actually published for this model, check its Hugging Face page before relying on this row.
Cheapest way to run Qwen3-8B-FP8-dynamic at its published (F8_E4M3) precision: 1× RTX 4070 Super, at $0.121/hr per GPU ($0.121/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
Qwen3-8B-FP8-dynamic: KV cache by context length
The KV cache is the memory the attention layers hold for every token of context, on top of the weights. It is not part of the flat 1.2x overhead in the table above, grows with context length and with every concurrent request, and is why a long-context deployment needs more VRAM than the table shows.
Grouped-query attention: every layer caches keys and values for a few shared KV heads. Cached per token: 147,456 bytes at 16-bit.
| Context | KV cache, one sequence |
|---|---|
| 4K | 0.56 GB |
| 32K | 4.5 GB |
| 40K (model maximum) | 5.6 GB |
Computed from the layer, head and window counts in the model's own configuration (RedHatAI/Qwen3-8B-FP8-dynamic), at 16 bits per cached value. Each concurrent sequence needs its own cache, on top of the weights.
Qwen3-8B-FP8-dynamic: common questions
Does Qwen3-8B-FP8-dynamic fit on a 12 GB GPU?
Yes. At FP8 (native) it needs 9.2 GB of VRAM, so a 12 GB card holds it with 2.8 GB to spare. An 8 GB card is not enough for it at FP8 (native).
Is Qwen3-8B-FP8-dynamic already quantized?
Yes. It is published in FP8, one byte per parameter, so the 9.2 GB figure is already a quantized footprint rather than a full-precision one. Only INT4 goes below it, at 4.6 GB. A GPU without FP8 tensor cores cannot run it as published, which is why cards here are matched on precision support and not on VRAM alone.
What is the least VRAM Qwen3-8B-FP8-dynamic can run in?
4.6 GB, at INT4 (quantized), which fits a 6 GB card, against 9.2 GB at FP8 (native). That row assumes an INT4 checkpoint has actually been published for this model, so check its Hugging Face page before planning around it.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More Qwen3 models
- Qwen3-8B-FP8 (8.2B, F8_E4M3)
- Qwen3-8B (8.2B, BF16)
- Qwen3-8B-FP8 (8.2B, F8_E4M3)
- Qwen3-8B-Base (8.2B, BF16)
- Qwen3Guard-Gen-8B (8.2B, BF16)
Lineage
Based on Qwen3-8B.
Alternatives at this size
Other models for text-generation within about a third of Qwen3-8B-FP8-dynamic's 8.2B parameters, from other model lines.
- Llama-3.1-8B-Instruct (8.0B, BF16)
- Mistral-7B-Instruct-v0.2 (7.2B, BF16)
- granite-4.1-8b (8.8B, BF16)
- DeepSeek-R1-0528-Qwen3-8B (8.2B, BF16)
- gemma-2-9b-it (9.2B, BF16)
More on Qwen3-8B-FP8-dynamic
Fits on a 12 GB GPU at FP8: every model that fits in 12 GB.
Fits on an 8 GB GPU at INT4: every model that fits in 8 GB.
Related reading: H100 pricing and specs, How much VRAM you need for LLMs, Serving LLMs with vLLM, vLLM vs TensorRT-LLM vs SGLang, and Best GPU for LLM inference.
