Gemma 3n E4B Instruct Preview (May '25)
May 2025
Gemma 3n E4B Instruct Preview (May '25)Compare inon
Benchmarks of Gemma 3n E4B Instruct Preview (May '25)4
BenchmarkByScoreRank
Overall1
Science1
Legacy · saturated or no longer reported2
No benchmark matches
—per task?
self-host onlyprice
—output speed?
32Kcontext
Ranking?
Strong at
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-6 Lunalow×3.7—$0.005Function calling only with none · Chat Completions+3
OpenMiMo v2.6 Prothinking×8—$0.13Pass back reasoning_content+3
BetterGranite 4.2 3BthinkingOpen+56%—$0.006BetterGemma 4 E4BthinkingOpen+53%—$0.008BetterGPT-5.6 Lunalow×3.6—$0.01ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-6 Lunalow×3.7—$0.005Function calling only with none · Chat Completions+3Granite 4.2 3BthinkingOpen+56%—$0.006Gemma 4 E4BthinkingOpen+53%—$0.008GPT-5.6 Lunalow×3.6—$0.01GPT OSS 20BhighOpen+54%—$0.011ModelScoreCostPer taskWhat changes
MiMo v2.6 ProthinkingOpen×8—$0.13Pass back reasoning_content+3
DeepSeek V4.1 Flash 1017maxOpen×7.8—$0.26Pass back reasoning_content · Chat Completions+4
GLM 5.3maxOpen×7.7—$2.01Reasoning cannot be disabled+3
Step 5 PreviewthinkingOpen×7.5—$1.03web_search not available · Chat Completions+2
Kimi K3maxOpen×7.5—$2.00Sampling parameters are fixed+3ModelScoreCostPer taskWhat changes
Gemini 4 Argonhigh×9—*$1.99Synchronous output stops at 65,536+2Gemini 3.8 Flashmedium×6.8—*$0.93Sampling parameters are ignored+1Gemini 3.6 Flashhigh×5.8—*$1.60Gemini 3.1 Pro Previewthinking×5.1—$1.30Cannot use both thinking controlsGemini 3.5 Flash Litehigh×3.8—$0.19Gemini 3 Flash Previewnon-thinking×3.1—$0.50 / $3Gemma 4 26B A4BthinkingOpen×2.9—$0.15 / $0.60Gemini 2.5 Prothinking×2.8—$0.33All Google models →Run it yourself
Fits onhardware needs not verified yet
LicenceGemma termscommercial use with conditions
Benchmarks
all categoriesChat & Knowledge
Humanity's Last Exam4.8#284
AA Intelligence Index 4.3.2≈5.8#301
GPQA Diamond27.8#304
Coding
LiveCodeBench · AA archived13.8#143
FAQ
How much does Gemma 3n E4B Instruct Preview (May '25) cost?›
There is no per-unit API price: self-host only.
Is Gemma 3n E4B Instruct Preview (May '25) free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Gemma 3n E4B Instruct Preview (May '25), and what is it best at?›
#282 of 334 in Chat & Knowledge, #142 of 152 in Coding.
Is Gemma 3n E4B Instruct Preview (May '25) better than Granite 4.1 3B?›
On AA Intelligence Index 4.3.2, Gemma 3n E4B Instruct Preview (May '25) scores 5.8 vs 5.9 — about the same.
What is the context window of Gemma 3n E4B Instruct Preview (May '25)?›
32K tokens.
When was Gemma 3n E4B Instruct Preview (May '25) released? Is there a newer version?›
Released 2025-05-20. It is the current version in its line.
Is Gemma 3n E4B Instruct Preview (May '25) open source?›
Yes, the weights are open (gemma licence).
Can I run Gemma 3n E4B Instruct Preview (May '25) locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Gemma 3n E4B Instruct Preview (May '25) read images?›
It accepts text and audio, video as input.
What is Gemma 3n E4B Instruct Preview (May '25)?›
Gemma 3n E4B Instruct Preview (May '25) is an open-weights language model from Google, released May 2025.
What are the best alternatives to Gemma 3n E4B Instruct Preview (May '25)?›
GPT-6 Luna (better), MiMo v2.6 Pro (open weights).