LangWatch
Sign Up

GoogleOpen weights model4B parameters

Gemma 3 4B IT benchmarks

All models
  • General rank
    #28of 28Leads 0 of 5 (0%)non-coding tasks
  • Cost
    Self-hostedSmall GPU, ~3.1 GBNo hosted price
  • LangWatch policy#5 / 42 tied
    22.4%Best task · PII left invs PII detectors

Comparison summary

Strongest
Safety and guardrails
Leads 0 of 5 (0%)

A language model (LLM): it reads a prompt and writes text back, so the prompt sets the task. Strongest at safety.

Specifications

Lab
Google
Weights
Open weights
Parameters
4B
Licence
Gemma Terms of Use
Type
small LLM

Overall standing

Share of non-coding tasks led · #28 of 28

  1. Claude Opus 5.585%
  2. OpenAI Decisions API67%
  3. Jev61%
  4. Eikos-27B60%
  5. GPT-6.1 Sol54%
  6. AutoJev-27B53%
  7. Claude Sonnet 5.546%
  8. Gemma 3 4B IT
    0%

Safety and guardrails

Share of tasks led · #14 of 46

  1. Gemma 4 31BLeads 3 of 4
  2. Perplexity PII TracerLeads 2 of 4
  3. Amazon ComprehendLeads 1 of 4
  4. Qwen3.8 27BLeads 1 of 5
  5. Gemma 4 12BLeads 0 of 4
  6. Ministral 3 14BLeads 0 of 5
  7. Gemma 4 26B A4BLeads 0 of 4
  8. Gemma 3 4B IT
    Leads 0 of 5
See the 5 tasks

Traces

Tier 6 of 19#13 / 46

PII left in, % · Lower is better

  1. Amazon Comprehend10.6, interval 9.4 to 11.9, ahead alone
  2. GLiNER Multi PII v115.3, interval 13.2 to 17.3, tier 2
  3. NVIDIA GLiNER PII19.7, interval 17.4 to 22.0, tier 3
  4. Gemma 4 26B A4B19.8, interval 15.8 to 24.2, tier 3
  5. Qwen3.8 27B20.3, interval 16.5 to 24.3, tier 3
  6. GLiNER2 Large20.4, interval 18.0 to 22.8, tier 3
  7. Gemma 4 12B23.3, interval 19.0 to 27.6, tier 3
  8. Gemma 3 4B IT
    43.9, interval 39.3 to 48.4, tier 6
Tested against 52 PII detectors

Business

Tier 4 of 18#9 / 46

PII left in, % · Lower is better

  1. Perplexity PII Tracer2.1, interval 1.2 to 3.0, tied for the lead
  2. Qwen3.8 27B2.4, interval 1.7 to 3.3, tied for the lead
  3. Gemma 4 31B2.9, interval 1.9 to 4.0, tied for the lead
  4. GLiNER Multi PII v14.1, interval 3.1 to 5.3, tier 2
  5. Ministral 3 14B5.4, interval 4.0 to 6.9, tier 2
  6. Gemma 4 12B5.8, interval 4.4 to 7.3, tier 3
  7. Amazon Comprehend5.9, interval 4.6 to 7.2, tier 3
  8. Gemma 3 4B IT
    10.7, interval 9.1 to 12.5, tier 4
Tested against 53 PII detectors

Multilingual

Tier 12 of 25#19 / 43

PII left in, % · Lower is better

  1. Perplexity PII Tracer2.9, interval 2.5 to 3.4, ahead alone
  2. Qwen3.8 27B4.2, interval 3.7 to 4.8, tier 2
  3. Gemma 4 31B4.5, interval 3.9 to 5.1, tier 2
  4. GLiNER Multi PII v15.7, interval 5.2 to 6.3, tier 3
  5. bardsai EU PII Multilang6.9, interval 6.3 to 7.5, tier 4
  6. Qwen3.6 35B A3B7.7, interval 7.0 to 8.4, tier 4
  7. Gemma 4 12B8.9, interval 8.1 to 9.7, tier 5
  8. Gemma 3 4B IT
    35.4, interval 33.9 to 36.7, tier 12
Tested against 53 PII detectors

Look-alikes

Tier 5 of 7#6 / 9

Look-alikes redacted, % · Lower is better

  1. Gemma 4 31B3.5, interval 2.1 to 5.1, ahead alone
  2. Qwen3.8 27B10.3, interval 7.9 to 12.9, tier 2
  3. NuExtract 319.8, interval 15.9 to 23.6, tier 3
  4. Qwen3.5 9B20.5, interval 16.8 to 24.3, tier 3
  5. Ministral 3 14B30.5, interval 26.5 to 34.8, tier 4
  6. NVIDIA GLiNER PII42.0, interval 38.2 to 46.1, tier 5
  7. Gemma 3 4B IT
    45.0, interval 41.0 to 49.0, tier 5
  8. Perplexity PII Tracer67.6, interval 63.7 to 71.6, tier 6
Tested against 53 PII detectors

LangWatch policy

Tier 4 of 13#5 / 42

PII left in, % · Lower is better

  1. Gemma 4 31B6.9, interval 2.8 to 11.3, ahead alone
  2. Ministral 3 14B10.2, interval 5.8 to 15.4, tier 2
  3. Gemma 4 12B16.5, interval 10.3 to 23.2, tier 3
  4. Qwen3.8 27B16.7, interval 9.8 to 25.0, tier 3
  5. Gemma 4 26B A4B20.5, interval 13.3 to 28.8, tier 4
  6. Qwen3.5 9B21.5, interval 13.4 to 30.6, tier 4
  7. Gemma 3 4B IT
    22.4, interval 14.3 to 32.1, tier 4
  8. Qwen3 4B Instruct 250729.6, interval 19.6 to 40.3, tier 5
Tested against 57 PII detectors

Cost

Self-hosted · GPU memory needed, among small LLMs · lower is better

  1. Qwen3 0.6B~1 GB
  2. Qwen3.5 0.8B~1.1 GB
  3. LFM2 1.2B~1.3 GB
  4. LFM2.5 1.2B Instruct~1.3 GB
  5. Gemma 3 1B IT~1.4 GB
  6. Llama 3.2 1B Instruct~1.4 GB
  7. GDPR Anonymization 0.5B~1.5 GB
  8. Gemma 3 4B IT
    ~3.1 GB
How we count

A model leads a task when it is in the task's leading tie tier. Each task counts inside its own benchmark, against that benchmark's models; no score is averaged. An area chart shows the models ranked on at least 2 of the area's tasks in a benchmark this model is in too. The overall standing counts every non-coding tasks and needs 5 for a rank. Latency is compared only on one hardware tier.

Gemma 3 4B ITOther modelsWhisker: 95% interval