Apertus-8B-Instruct-2509
Where it runs 8K context · ≥ 5 t/s floor
| Setup | Best plan | Memory | Generate |
|---|---|---|---|
| 8 GB GPU + 32 GB RAM | Q5_K_Mllama.cpp CUDAFull GPU | Fits | ~26–37 t/s Estimated |
| 12 GB GPU + 32 GB RAM | Q8_0llama.cpp CUDAFull GPU | Fits | ~30–32 t/s Estimated |
| 16 GB GPU + 64 GB RAM | Q8_0llama.cpp CUDAFull GPU | Fits | ~63–77 t/s Estimated |
| 24 GB GPU + 64 GB RAM | FP16llama.cpp CUDAFull GPU | Fits | ~35–43 t/s Estimated |
| 32 GB GPU + 128 GB RAM | FP16llama.cpp CUDAFull GPU | Fits | ~54–87 t/s Estimated |
| 2× 24 GB GPUs + 128 GB RAM | FP16llama.cpp CUDAFull GPU | Fits | ~32–41 t/s Estimated |
| 96 GB workstation GPU + 256 GB RAM | FP16llama.cpp CUDAFull GPU | Fits | ~56–87 t/s Estimated |
| Mac, 24 GB | FP16llama.cpp MetalUnified memory | Fits | ~12–14 t/s Estimated |
| Mac, 48 GB | FP16llama.cpp MetalUnified memory | Fits | ~12–14 t/s Estimated |
| Mac, 128 GB | FP16llama.cpp MetalUnified memory | Fits | ~19–29 t/s Estimated |
| Mac, 512 GB | FP16llama.cpp MetalUnified memory | Fits | ~21–32 t/s Estimated |
| DGX Spark 128 GB | FP16llama.cpp CUDAUnified memory | Fits | ~12–14 t/s Estimated |
| Ryzen AI Max+ 395, 128 GB | FP16llama.cpp VulkanUnified memory | Fits | ~8.4–12 t/s Estimated |
| CPU only, 64 GB DDR5-5600 dual channel | Q4_K_Mllama.cpp CPUCPU only | Fits | ~3.6–4.3 t/s Estimated extrapolated |
Same plan engine as the planner, with its default reserves and RAM. How plans are chosen.
Files exact Hugging Face bytes
| Artifact | Size | Files | Revision | Source |
|---|---|---|---|---|
| FP16 mradermacher/Apertus-8B-Instruct-2509-GGUF | 15.01 GiB | 1 | 3ff96561e4 | HF files |
| Q8_0 bartowski/swiss-ai_Apertus-8B-Instruct-2509-GGUF | 7.98 GiB | 1 | d2ffbcc459 | HF files |
| Q5_K_M bartowski/swiss-ai_Apertus-8B-Instruct-2509-GGUF | 5.41 GiB | 1 | d2ffbcc459 | HF files |
| Q4_K_M bartowski/swiss-ai_Apertus-8B-Instruct-2509-GGUF | 4.71 GiB | 1 | d2ffbcc459 | HF files |
| MLX 8-bit 8-bit mlx-community/Apertus-8B-Instruct-2509-8bit | 7.97 GiB | 2 | fa890d8c0f | HF files |
| MLX 4-bit 4-bit mlx-community/Apertus-8B-Instruct-2509-4bit | 4.22 GiB | 1 | c5e6ae2e52 | HF files |
| Native safetensors BF16 swiss-ai/Apertus-8B-Instruct-2509 | 15.00 GiB | 4 | b946d40447 | HF files |
Not found in the representative quant search (not proven unavailable): EXL3, AWQ, GPTQ. Complete selected file sets; projectors, MTP/draft and lookup files are listed separately and budgeted only by plans that use them. File bytes are not allocated memory.
Quality
| AA evaluated mode | Index | LiveCodeBench | Terminal-Bench 2.1 | GPQA | HLE | Source |
|---|---|---|---|---|---|---|
| Apertus 8B Instruct AA estimate | 4.8 | 0.0% | — | 25.6% | 5.0% | source |
Arena text rating
Unknown
Publisher-reported benchmarks
None extracted from the model card.
AA and Arena describe the hosted model in the stated mode, not each quantized file; scores are never inherited by base, pruned, abliterated or quantized variants and never averaged together.
How to run it
No model-specific recipe in the research. The planner binds the backend’s documented command template (llama.cpp, Ollama, LM Studio, MLX, vLLM) to your plan: open a plan’s details.
Recipes are source examples (settings are the publisher’s or author’s, not measured optima) and were not executed by LocalFit. A downloadable file does not prove your backend build supports the architecture.
Reported speeds 0
No reported measurement yet. Submit one.