Google DeepMind

Gemma 3 4B IT

Type
language model
Context
131K tokens
Max output
16K
Released
12 March 2025
Knowledge cutoff
August 2024
API string
google/gemma-3-4b-it
Weights
open

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
Amazon Bedrock $0.04 $0.08 31 Jul 2026
DeepInfra $0.04 $0.08 31 Jul 2026
Kilo Code $0.04 $0.08 31 Jul 2026
OpenRouter $0.05 $0.1 31 Jul 2026
NanoGPT $0.201 $0.201 31 Jul 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
Arena Score in English 1,300.01 1,288–1,312 — —
Arena Score in Russian 1,294.12 1,264–1,325 — —
Arena Score, overall 1,290.88 1,282–1,300 — —
Arena Score, long queries 1,273.61 1,247–1,301 — —
Arena Score, creative writing 1,271.03 1,249–1,293 — —
Arena Score, multi-turn dialogue 1,255.13 1,232–1,279 — —
Arena Score, hard prompts 1,252.74 1,236–1,270 — —
Arena Score, instruction following 1,239.54 1,223–1,256 — —
Arena Score, mathematics 1,238.96 1,211–1,267 — —
Arena Score, programming 1,230.79 1,207–1,254 — —
Arena Score, expert questions 1,222.75 1,182–1,264 — —