Google: Gemma 3n 4B

Gemma 3n 4B by Google is designed for text-based tasks with a context length of 32,768 tokens; it does not support tools or reasoning, nor does it generate structured output. Gemma excels in handling plain text inputs but lacks the capability to integrate external tools or perform complex reasoning processes. This model should be considered by those on a budget looking for basic text processing tasks. With pricing at $0.06 per million input tokens and $0.12 per million output tokens, it offers competitive cost-effectiveness. While its blended benchmark score of 0.4 across two benchmarks is modest, it provides adequate performance for routine, straightforward applications where advanced reasoning or tool integration are not required.