Gemma 4 26B A4B IT Assistant — Hardware Requirements & GPU Compatibility
Chat
Gemma 4 26B A4B IT Assistant is a 26B-parameter open language model from Google in the Gemma 4 family. It supports a context window of up to 262,144 tokens. At Q4_K_M it needs about 15.92 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 12 GB (Q2_K · 64K)Best on 24 GB (Q4_K_M · ~253 tok/s · 256K)Macs from 16 GB (Q2_K · ~75 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Gemma 4 26B A4B IT Assistant
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run hf.co/AtomicChat/gemma-4-26B-A4B-it-assistant-GGUF:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 8.0 GB free · 16K context
Download: gemma-4-26B-A4B-it-assistant.Q4_K_M.gguf0.33 GBAtomicChat
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3090 (24 GB) (8 GB headroom) · Q4_K_M
- Generation speed
- ~236 tok/s
- generation speed
- Cost per 1M output tokens
- $0.20
- per 1M output tokens
How Much VRAM Does Gemma 4 26B A4B IT Assistant Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 11.4 GB | 11.6 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 256K context · ~346 tok/s | |
| 13.0 GB | 13.2 GB | 3-bit medium — 3-bit medium quantization | Fits: 256K context · ~306 tok/s | |
| 14.9 GB | 15.2 GB | 4-bit small — 4-bit small quantization | Fits: 256K context · ~268 tok/s | |
| 15.9 GB | 16.2 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 256K context · ~253 tok/s | |
| 18.8 GB | 19.1 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 256K context · ~215 tok/s | |
| 21.8 GB | 22 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 256K context · ~188 tok/s | |
| 26.3 GB | 26.6 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Does not fit: needs 26.4 GB | |
| 52.3 GB | 52.6 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 52.4 GB |
Selected Q4_K_M
4-bit medium quantization16.0GBat 16K context
- Weights
- 15.6 GB
- KV cache
- 0.1 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 8.0 GB free · ~253 tok/s
≥50 tok/s needs ≥190 GB/s · RTX 5090 ~431, RTX 3090 Ti ~253
Which Hardware Can Run Gemma 4 26B A4B IT Assistant?
Q4_K_M · 15.9 GBGemma 4 26B A4B IT Assistant (Q4_K_M) requires 15.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 21+ GB is recommended. Using the full 262K context window can add up to 2.1 GB, bringing total usage to 18.1 GB. 26 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 5080. 122 devices with unified memory can run Gemma 4 26B A4B IT Assistant, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Mini M5 Pro (24 GB).
20 GB
fits · 4.0 GB free · up to 256K ctx
1 GPU
RX 7900 XT~169Simulate output on RX 7900 XT32–40 GB
fits · 16.0 GB free · up to 256K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~431Simulate output on RTX 5090M5 Max36 GB~119Simulate output on Mac Studio M5 Max (36 GB)M532 GB~42Simulate output on MacBook Air 13" M5 (32 GB)Orin~54Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~35Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to Gemma 4 26B A4B IT Assistant
Gemma 4 family · 35 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Qwen3.8 27BAlibaba | 122.0 | 17.9 GB | ~38 | same size |
| Qwen3.6 27BAlibaba | 111.2 | 17.9 GB | ~38 | same size |
| Qwen3.5 27BAlibaba | 103.8 | 17.9 GB | ~38 | same size |
| Ariarhymes-ai | — | 20.2 GB | ~168 | same size |
| JiRackDeltaNet 27BCMSManhattan | — | 17.6 GB | ~38 | same size |
| Fanar 2 27B InstructQCRI | — | 19.3 GB | ~37 | same size |
Fine-tunes & variants(1)
Specifications
- Publisher
- Family
- Gemma 4
- Parameters
- 26B
- Active parameters
- 4B per token
- Architecture
- Gemma4AssistantForCausalLM
- Context Length
- 262,144 tokens
- Vocabulary Size
- 262,144
- Release Date
- 2026-04-23
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Gemma 4 26B A4B IT Assistant need?
Gemma 4 26B A4B IT Assistant requires 15.9 GB of VRAM at Q4_K_M, or 52.3 GB at BF16. Full 262K context adds up to 2.1 GB (18.1 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 26B × 4.8 bits ÷ 8 = 15.6 GB
KV Cache + Overhead ≈ 0.3 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 2.5 GB (at full 262K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 4090 run Gemma 4 26B A4B IT Assistant?
Yes, at Q6_K (21.8 GB) or lower. Higher quantizations like Q8_0 (26.3 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for Gemma 4 26B A4B IT Assistant?
For Gemma 4 26B A4B IT Assistant, Q4_K_M (15.9 GB) offers the best balance of quality and VRAM usage. Q5_K_M (18.8 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 11.4 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Gemma 4 26B A4B IT Assistant on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Gemma 4 26B A4B IT Assistant, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 15.9 GB of usable unified memory (RAM minus macOS overhead), from Mac Mini M5 Pro (24 GB).
Can I run Gemma 4 26B A4B IT Assistant locally?
Yes — Gemma 4 26B A4B IT Assistant can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Gemma 4 26B A4B IT Assistant?
At Q4_K_M, Gemma 4 26B A4B IT Assistant can reach ~763 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~253 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)
Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.
Example: NVIDIA B200 → 2.4 GB active ÷ (8000 × 0.65) = 0.471 ms, plus 4 layers × 0.055 ms = 0.220 ms, so 1000 ÷ 0.691 ms = ~1447 tok/s
Estimated speed at Q4_K_M (15.9 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Gemma 4 26B A4B IT Assistant?
At Q4_K_M, the download is about 15.60 GB. The full-precision BF16 version is 52.00 GB. The smallest option (Q2_K) is 11.05 GB.
Which GPUs can run Gemma 4 26B A4B IT Assistant?
26 consumer GPUs can run Gemma 4 26B A4B IT Assistant at Q4_K_M. Top options include AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090, NVIDIA GeForce RTX 3090 Ti, AMD Radeon RX 6800. 7 GPUs have plenty of headroom for comfortable inference.
Which devices can run Gemma 4 26B A4B IT Assistant?
122 devices with unified memory can run Gemma 4 26B A4B IT Assistant at Q4_K_M (15.9 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (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.