Skip to content
OSMEXO

Gemma 3 4B IT

google/gemma-3-4b-it
  • Google
  • Family · Gemma 3
  • Open weights
  • Text · Image → Text
  • JSON · Vision
Not on OSMEXO yet
Spec & list price
Context128K
Max output16K
Input / 1M$0.05
Output / 1M$0.102× input
Usage
Tokens / day1.1B▲ 260% 7d
Usage rank#191
Requests / day5.1M
Tool calls / req0.00
Per request · 30 days
Avg input2370.0× median
Avg output440.1× median
Cache hit6.9%0.1× median
Tool-call failures100%52× median

Consumption

500.0M1.0B1.5B2 Oct16 Oct31 Oct
Peak 1.5B 30 OctLow 69.3M 10 OctRank today #191

News & trends

2 mentions this weekGemma 3 family · videos, posts, articles · 30 days
News · latest from Google

Benchmarks

Benchmark data: Artificial Analysis
Output speed#243/33347 tok/s1.4× slower than median
Fastest provider50 tok/sbest of 2 providers
First token#480.45 s2.3× quicker than median
Family progress
Intelligence Index
3.544.555.5AprMayJunJulAug
  • Family releases5.1

Technical details

Tokenizer
Gemini
Knowledge cutoff
2024-08-31
License
gemma
Parameters
4.3B
Moderation
none
Downloads · likes
955.8K · 1.6K
Request parameters 15
  • response_format
  • structured_outputs
  • temperature
  • top_p
  • top_k
  • min_p
  • max_tokens
  • stop
  • seed
  • frequency_penalty
+5 more
  • logit_bias
  • logprobs
  • presence_penalty
  • repetition_penalty
  • top_logprobs

When to use

Best for

  • On-device or edge text and image assistants
  • Low-cost bulk classification and tagging
  • Self-hosted chat on modest hardware
  • Image captioning and simple visual Q&A
  • Summarizing long inputs cheaply

Think twice for

  • Hardest reasoning and coding tasks
  • Tool-calling agents
  • Image or audio generation

Research papers

Same family

Gemma 3

Want Gemma 3 4B IT on OSMEXO?

Sign in to request it. Requests are reviewed; a request does not guarantee availability.