Darwin 36B Opus — Hardware Requirements & GPU Compatibility
ChatReasoning
Darwin 36B Opus is a 34.7B-parameter open language model from FINAL-Bench. It supports a context window of up to 262,144 tokens. At BF16 it needs about 69.71 GB of VRAM — see which GPUs and Macs can run it below.
2.0K downloads
71 likesSpecifications
- Publisher
- FINAL-Bench
- Parameters
- 34.7B
- Architecture
- Qwen3_5MoeForCausalLM
- Context Length
- 262,144 tokens
- Vocabulary Size
- 248,320
- Release Date
- 2026-06-03
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Darwin 36B Opus Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 69.7 GB | 80.4 GB | 69.32 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Darwin 36B Opus?
BF16 · 69.7 GBDarwin 36B Opus (BF16) requires 69.7 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 91+ GB is recommended. Using the full 262K context window can add up to 10.7 GB, bringing total usage to 80.4 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Darwin 36B Opus?
BF16 · 69.7 GB5 devices with unified memory can run Darwin 36B Opus, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Related Models
Frequently Asked Questions
- How much VRAM does Darwin 36B Opus need?
Darwin 36B Opus requires 69.7 GB of VRAM at BF16. Full 262K context adds up to 10.7 GB (80.4 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 34.7B × 16 bits ÷ 8 = 69.3 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 11.1 GB (at full 262K context)
VRAM usage by quantization
BF1669.7 GBBF16 + full context80.4 GB- Can NVIDIA GeForce RTX 5090 run Darwin 36B Opus?
No — Darwin 36B Opus requires at least 69.7 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run Darwin 36B Opus on a Mac?
Darwin 36B Opus requires at least 69.7 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Darwin 36B Opus locally?
Yes — Darwin 36B Opus can run locally on consumer hardware. At BF16 quantization it needs 69.7 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Darwin 36B Opus?
At BF16, Darwin 36B Opus can reach ~42 tok/s on AMD Instinct MI300X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 69.7 × 0.55 = ~42 tok/s
Estimated speed at BF16 (69.7 GB)
~42 tok/s~31 tok/s~26 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Darwin 36B Opus?
At BF16, the download is about 69.32 GB.
- Which GPUs can run Darwin 36B Opus?
No single consumer GPU has enough VRAM to run Darwin 36B Opus at BF16 (69.7 GB). Multi-GPU or professional hardware is required.
- Which devices can run Darwin 36B Opus?
5 devices with unified memory can run Darwin 36B Opus at BF16 (69.7 GB), including Mac Pro M2 Ultra (192 GB), Mac Studio M2 Ultra (192 GB), Mac Studio M4 Max (128 GB), NVIDIA DGX A100 640GB. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.