google

Google: Gemma 3 4B

The Google Gemma 3 4B is an AI model that supports two input modalities: text and image. It has a large context length of 131072 tokens, allowing for complex inputs and outputs. This model may be worth considering for those who require a mix of text and image processing capabilities, particularly at its price point. With a blended benchmark score of 14.5 across three independent benchmarks, it holds a middle ground in terms of performance. However, users should also consider the cost per million input tokens ($0.05) and output tokens ($0.1), which may impact their decision depending on their specific needs.

Open weights 4.3B parameters, about 2.4 GiB at Q4_K_M: fits comfortably on 17 of 17 consumer graphics cards in the PicksByCard catalog. See which cards run it →
Quality Score
81/100
price + capability + benchmarks
Input Price
$0.05
per 1M tokens · checked 2026-09-04
Output Price
$0.10
per 1M tokens · checked 2026-09-04
Context Window
131,072
tokens

Benchmark results

Independent, published benchmarks. Blended score 14.5 across 3 benchmarks, last refreshed 2026-09-04. How scoring works →

BenchmarkMeasuresScore
AI Index broad capability composite 1.6
AI Index Coding software engineering tasks 4.4
EQ-Bench emotional understanding in dialogue 42.4
Model ID
google/gemma-3-4b-it
Vendor
google
Released
March 2025
Tokenizer
Gemini
Input Modalities
text, image
Output Modalities
text
Max Output
16,384 tokens
Tool Calling
not supported
Structured Output
✓ supported
Reasoning Mode
not supported
Vision
✓ accepts images
Audio
no
Moderated
no

What it costs in practice

Computed from the current $0.05/M input and $0.10/M output rates. Run your own numbers →

JobTokensCost
Summarize a 50-page report 30k in / 1.5k out under $0.01
Classify 1,000 customer emails 500k in / 50k out $0.03
A month of a busy support chatbot 5M in / 2M out $0.45

Similar models

Quick answers

How much does Google: Gemma 3 4B cost?
$0.05 per million input tokens and $0.10 per million output tokens.
What is Google: Gemma 3 4B's context window?
131,072 tokens, roughly 196 pages of text in a single request.
Does Google: Gemma 3 4B support tool calling?
It supports structured output.
Can Google: Gemma 3 4B process images?
Yes, it accepts image input.