What is Gemma 4 26B A4B good for?

Self-hosted text and image work at near the speed of a 4B dense model, from a mixture of experts that activates 3.8B parameters per token. Apache 2.0 weights, on a single consumer GPU.

When is Gemma 4 26B A4B not a good fit?

Audio or video input, and hosted tools like web search or code execution. Memory follows the full 25.2B parameters, not the 3.8B active, and no maximum output is published.

What is the cheapest way to run Gemma 4 26B A4B?

Hosted, unless you push serious volume. Cloudflare charges $0.10 in / $0.30 out per 1M tokens. The cheapest rental that fits is 2x RTX 4070 at $144 a month, which costs the same as about 1B tokens a month on that API.

Can I self-host Gemma 4 26B A4B?

Yes. The weights are Apache 2.0 licensed. At 4-bit it needs about 20 GB of GPU memory, which starts at roughly $144 a month on the cheapest rental that fits.

Open weights · Apache 2.0

Gemma 4 26B A4B

A 25.2B-parameter MoE (3.8B active, 128 experts) in Google's Gemma 4 family, running near the speed of a 4B dense model. Apache 2.0 license.

Jump to: Specs Pricing Self-hosting Similar models FAQ More from Google Cloud

Visit website Compare vs other LLMs →

Cheapest via Cloudflare

Input: $0.10 / 1M tokens
Output: $0.30 / 1M tokens

3 providers

Key Specifications

Context window: 256K tokens
Knowledge cutoff: Jan 2025
Released: Apr 2026
Parameters: 25.2B, 3.8B active
Inputs: Text, image

Capabilities Show details: Reasoning Thinks before it answers, always on or as a switchable mode. Function calling Connect to external tools, APIs, and systems. Structured output Return responses in structured formats like JSON.

Hosted API pricing

Input tokens M Output tokens M

Provider	Input / 1M tokens	Output / 1M tokens	Cost at 10M in + 2M out
Cloudflare Cheapest	$0.10	$0.30	$1.60	View
Novita	$0.13	$0.40	$2.10	View
Google Cloud Creator	Open weights, no hosted price			View

Heads up: Base-tier, on-demand rates per 1M tokens; cached, batch and long-context tiers excluded. A provider may serve a shorter context or a quantized build than the creator's release. Verify before provisioning. More on how we price.

Estimated cost to self-host

Gemma 4 26B A4B needs about 20 GB of GPU memory at 4-bit with 32K context. The cheapest rental that fits is 2× RTX 4070 at about $144 a month.

That costs the same as roughly 1B tokens a month on Cloudflare's API. Self-hosting is more expensive below that volume.

Precision	Cheapest, 32K context	Cheapest, full 256K context
4-bitINT4 / FP4	2× RTX 4070 20 GB VRAM · $144/mo	4× RTX 4060 Ti 48 GB VRAM · $317/mo
8-bitFP8 / INT8	4× RTX 3060 35 GB VRAM · $173/mo	8× RTX 4070 59 GB VRAM · $576/mo
16-bitFP16 / BF16	8× RTX 3060 60 GB VRAM · $346/mo	8× RTX 3090 116 GB VRAM · $979/mo

Estimates based on median on-demand rates for Nvidia GPUs. Memory is weights plus KV cache for one request, using FP8 KV cache where supported and FP16/BF16 otherwise. KV cache assumed at 250 KB per token, a conservative default. Break-even assumes a 5:1 input-to-output ratio. No guarantee of runtime support, usable performance, or that a matching quantized build exists. How we estimate costs.

Similarly priced models

The models nearest Gemma 4 26B A4B by blended rate, each at its own cheapest provider.

Model	Blended / 1M	Input / 1M	Output / 1M	Context	Cutoff	vs Gemma 4 26B A4B
Ministral 3 3B Mistral	$0.10	$0.10	$0.10	256K		−25%
DeepSeek V4 Flash DeepSeek	$0.1083	$0.08	$0.25	1M		−19%
GLM-5.3-Flash Z.AI	$0.1083	$0.08	$0.25	1M		−19%
GPT-5 Nano OpenAI	$0.1083	$0.05	$0.40	400K	May 2024	−19%
Gemma 3 27B Instruct Google Cloud	$0.1333	$0.10	$0.30	128K	Aug 2024	0%
Gemma 4 26B A4B This model Google Cloud	$0.1333	$0.10	$0.30	256K	Jan 2025
Qwen3-32B FP8 Alibaba Cloud	$0.1333	$0.10	$0.30	33K		0%
Gemini 2.5 Flash-Lite Google Cloud	$0.15	$0.10	$0.40	1M	Jan 2025	+13%
GPT-4.1 Nano OpenAI	$0.15	$0.10	$0.40	1M	Jun 2024	+13%
Ministral 3 8B Mistral	$0.15	$0.15	$0.15	256K		+13%
Qwen3.5-Flash Alibaba Cloud	$0.15	$0.10	$0.40	1M		+13%

Prices are USD per 1M tokens at each model's cheapest listed provider. Blended is the cost of 10M input plus 2M output tokens, spread over the 12M.

Runpod

Our sponsor

Spin Up a GPU in Seconds - No Setup Required

Deploy AI Models Globally in 30+ Regions

Scale from 0 to 1000 GPUs on Demand

PyTorch, Tensorflow & Docker Ready

Learn more

Frequently Asked Questions

What is Gemma 4 26B A4B good for?: Self-hosted text and image work at near the speed of a 4B dense model, from a mixture of experts that activates 3.8B parameters per token. Apache 2.0 weights, on a single consumer GPU.
When is Gemma 4 26B A4B not a good fit?: Audio or video input, and hosted tools like web search or code execution. Memory follows the full 25.2B parameters, not the 3.8B active, and no maximum output is published.
What is the cheapest way to run Gemma 4 26B A4B?: Hosted, unless you push serious volume. Cloudflare charges $0.10 in / $0.30 out per 1M tokens. The cheapest rental that fits is 2x RTX 4070 at $144 a month, which costs the same as about 1B tokens a month on that API.
Can I self-host Gemma 4 26B A4B?: Yes. The weights are Apache 2.0 licensed. At 4-bit it needs about 20 GB of GPU memory, which starts at roughly $144 a month on the cheapest rental that fits.

More from Google Cloud

Model	Context	Cutoff	Input / 1M	Output / 1M
Gemma 3 27B Instruct Open weights	128K	Aug 2024	$0.10	$0.30
Gemini 2.5 Flash-Lite	1M	Jan 2025	$0.10	$0.40
Gemma 4 31B Open weights	256K	Jan 2025	$0.13	$0.37
Gemma 3 12B Instruct Open weights	128K	Aug 2024	$0.05	$0.10
Gemini 3.1 Flash-Lite	1M	Jan 2025	$0.25	$1.50
Gemini 3.5 Flash-Lite	1M	Mar 2026	$0.30	$2.50
Gemini 2.5 Flash	1M	Jan 2025	$0.30	$2.50
Gemini 3 Flash Preview	1M	Jan 2025	$0.50	$3.00