Mellum2 12B A2.5B Thinking — Hardware Requirements & GPU Compatibility
Chat
Mellum2 12B A2.5B Thinking is a 12.1B-parameter open language model from JetBrains in the Mellum family. It supports a context window of up to 131,072 tokens. At Q4_K_M it needs about 7.66 GB of VRAM — see which GPUs and Macs can run it below.
5.2K downloads
320 likesSpecifications
- Publisher
- JetBrains
- Family
- Mellum
- Parameters
- 12.1B
- Architecture
- MellumForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 98,304
- Release Date
- 2026-05-26
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Mellum2 12B A2.5B Thinking Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 5.5 GB | 9.7 GB | 5.16 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 5.7 GB | 9.8 GB | 5.32 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 6.3 GB | 10.4 GB | 5.92 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 6.4 GB | 10.6 GB | 6.07 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 7.7 GB | 11.8 GB | 7.29 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 9.0 GB | 13.2 GB | 8.66 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 10.4 GB | 14.6 GB | 10.02 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 12.5 GB | 16.7 GB | 12.15 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run Mellum2 12B A2.5B Thinking?
Q4_K_M · 7.7 GBMellum2 12B A2.5B Thinking (Q4_K_M) requires 7.7 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 10+ GB is recommended. Using the full 131K context window can add up to 4.2 GB, bringing total usage to 11.8 GB. 50 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3080.
Runs great
— Plenty of headroom
Decent
— Enough VRAM, may be tightWhich Devices Can Run Mellum2 12B A2.5B Thinking?
Q4_K_M · 7.7 GB55 devices with unified memory can run Mellum2 12B A2.5B Thinking, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, iPad Pro M5 13" (16 GB).
Runs great
— Plenty of headroomWhere to Download Mellum2 12B A2.5B Thinking
Community quantizations of this model — GGUF for llama.cpp, Ollama, and LM Studio, plus AWQ/MLX variants where available.
Related Models
Frequently Asked Questions
- How much VRAM does Mellum2 12B A2.5B Thinking need?
Mellum2 12B A2.5B Thinking requires 7.7 GB of VRAM at Q4_K_M, or 24.7 GB at BF16. Full 131K context adds up to 4.2 GB (11.8 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 12.1B × 4.8 bits ÷ 8 = 7.3 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 4.5 GB (at full 131K context)
VRAM usage by quantization
Q4_K_M7.7 GBQ4_K_M + full context11.8 GB- Can NVIDIA GeForce RTX 4090 run Mellum2 12B A2.5B Thinking?
Yes, at Q8_0 (12.5 GB) or lower. Higher quantizations like BF16 (24.7 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
- What's the best quantization for Mellum2 12B A2.5B Thinking?
For Mellum2 12B A2.5B Thinking, Q4_K_M (7.7 GB) offers the best balance of quality and VRAM usage. Q4_K_L (7.8 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XS at 4.0 GB.
VRAM requirement by quantization
IQ2_XS4.0 GBQ3_K_S5.7 GBIQ4_XS6.9 GBQ4_K_M ★7.7 GBQ5_K_S8.7 GBBF1624.7 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Mellum2 12B A2.5B Thinking on a Mac?
Mellum2 12B A2.5B Thinking requires at least 4.0 GB at IQ2_XS, 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 Mellum2 12B A2.5B Thinking locally?
Yes — Mellum2 12B A2.5B Thinking can run locally on consumer hardware. At Q4_K_M quantization it needs 7.7 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Mellum2 12B A2.5B Thinking?
At Q4_K_M, Mellum2 12B A2.5B Thinking can reach ~574 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~86 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 7.7 × 0.65 = ~679 tok/s
Estimated speed at Q4_K_M (7.7 GB)
~679 tok/s~86 tok/s~679 tok/s~574 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Mellum2 12B A2.5B Thinking?
At Q4_K_M, the download is about 7.29 GB. The full-precision BF16 version is 24.30 GB. The smallest option (IQ2_XS) is 3.64 GB.
- Which GPUs can run Mellum2 12B A2.5B Thinking?
50 consumer GPUs can run Mellum2 12B A2.5B Thinking at Q4_K_M (7.7 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT, AMD Radeon RX 7600. 35 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Mellum2 12B A2.5B Thinking?
59 devices with unified memory can run Mellum2 12B A2.5B Thinking at Q4_K_M (7.7 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Apple iPhone 17 Pro, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.