Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d263Policy30d65, 14 pendingSources1,013

Gemma 3 4B

GoogleReleased Mar 12, 2025Current

Score basis

Gemma 3 4B is behind the leaders in multimodal tasks. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multilingual tasks, instruction following, or factuality.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Limited1 capability
  1. Multimodal−51.2%2 of 6
Not rated9 capabilities

Too few results yet to rate Reasoning, Coding, Agentic, Safety, Long Context, Math, Multilingual, Instruction Following or Factuality.

Price

$0.05input$0.10outputper million tokens

From deepinfra · 2 providers tracked · All prices

Cheaper than 95% of 331 priced models · 3:1 input-to-output blend, log scale

Evidence

61results on52benchmarks

  • 5 independently verified
  • 18 aggregator
  • 18 vendor-reported
  • 20 cross-referenced

From 7 sources · latest Oct 8, 2026 · How verification works

API features

Structured outputs

As listed by OpenRouter

Research

23 papers reference Gemma 3 4B

Gemma 3 4B benchmark results

61 results on 52 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Multimodal

LimitedFull multimodal ranking

51.2% behind the leader2 of 6 ranked benchmarks measured

Reasoning

Not enough dataFull reasoning ranking

0 of 6 ranked benchmarks measured

Show 3 more reasoning resultsHide 3 reasoning results

Coding

Not enough dataFull coding ranking

0 of 10 ranked benchmarks measured

Show 1 more coding resultHide 1 coding result

Agentic

Not enough dataFull agentic ranking

0 of 7 ranked benchmarks measured

Long Context

Not enough dataFull long context ranking

0 of 3 ranked benchmarks measured

Instruction Following

Not enough dataFull instruction following ranking

0 of 3 ranked benchmarks measured

Factuality

Not enough dataFull factuality ranking

0 of 4 ranked benchmarks measured

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 34 more resultsHide 34 results

Gemma 3 4B: common questions

Who makes Gemma 3 4B?

Gemma 3 4B is made by Google.

When was Gemma 3 4B released?

Gemma 3 4B was released on Mar 12, 2025, according to Artificial Analysis.

What is Gemma 3 4B good at?

Gemma 3 4B is behind the leaders in multimodal tasks. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multilingual tasks, instruction following, or factuality.

How much does Gemma 3 4B cost?

Gemma 3 4B costs $0.05 per million input tokens and $0.10 per million output tokens, according to deepinfra. We track its price at 2 providers. At a mix of three input tokens to one output token, it is cheaper than 95% of the 331 priced models we track.

How many benchmarks has Gemma 3 4B been tested on?

We track 61 results for Gemma 3 4B on 52 benchmarks from 7 sources, 5 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does Gemma 3 4B support?

OpenRouter lists structured outputs for Gemma 3 4B.

About this record

Where Gemma 3 4B's numbers come from, and every name it appears under.

Tracked since
May 15, 2026
Newest source mention
Aug 9, 2026

Where the results come from

Verification: 61 scores · 5 independently verified · 18 aggregator-attributed · 20 vendor cross-reference · 18 vendor-reported. How these tiers are assigned

From 7 sources on 4 sites. Hugging Face supplies 21 of them; the 5 independently verified results come from 2 sites. Bars are coloured by trust tier.

  • huggingface.co21
  • api.llm-stats.com18
  • artificialanalysis.ai18
  • raw.githubusercontent.com4

Also known as

Gemma 3 4B InstructGemma 3 4B ITgemma3-4b