Novita

google/gemma-3-12b-it

google/gemma-3-12b-it
Context

128K

Input

text, image

Output

text

Tool calling

Supported

About this model

Gemma 3 12B Instruct is a 12-billion-parameter multimodal model from Google, built from the same research as the Gemini models. It is instruction-tuned for dialogue and excels at text generation and image understanding tasks like question answering, summarization, and reasoning.

Compare Text Models

Input modalities

text

Accepted as model input

image

Accepted as model input

Providers

Available routing options for this model through OneInfer.

Novita
Available

Input tokens

$0.050 / 1M tokens

Output tokens

$0.100 / 1M tokens

Routing

OneInfer optimized

Pricing

Current OneInfer pricing for this model.

UsagePrice
Input tokens$0.050 / 1M tokens
Output tokens$0.100 / 1M tokens

Performance

Published evaluation results associated with this model.

Reasoning and Factuality

HellaSwag (10-shot)84.2
BoolQ (0-shot)78.8
ARC-Challenge (25-shot)68.9
DROP (3-shot, F1)72.2

STEM and Code

MMLU (5-shot)74.5
GSM8K (5-shot, maj@1)71
HumanEval (pass@1)45.7
MBPP (3-shot)60.4

Multimodal Understanding

DocVQA (val)82.3
TextVQA (val)66.5
AI2D75.2
ChartQA60.9

API example

Refer Documentation
curl https://api.oneinfer.ai/v1/ula/chat/completions \
  -H "Authorization: Bearer $ONEINFER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemma-3-12b-it",
    "messages": [
      { "role": "user", "content": "Hello!" }
    ]
  }'