Skip to main content

​
Overview

Google provides the Gemini family of multimodal models with strong capabilities in reasoning, vision, and grounding. Upsonic supports both the Gemini API and Vertex AI. Model Class: GoogleModel Provider Options:
  • google-gla: Gemini API (generativelanguage.googleapis.com)
  • google-vertex: Vertex AI API (requires GCP setup)

​
Authentication

export GOOGLE_API_KEY="AIza..."
# OR (legacy)
export GEMINI_API_KEY="AIza..."

# For Vertex AI
export GOOGLE_CLOUD_PROJECT="your-project-id"
export GOOGLE_CLOUD_LOCATION="us-central1"  # Optional

​
Examples

from upsonic import Agent, Task
from upsonic.models.google import GoogleModel

model = GoogleModel(model_name="gemini-2.5-flash", provider="google-gla")
agent = Agent(model=model)

task = Task("Hello, how are you?")
result = agent.do(task)
print(result.output)

​
Parameters

ParameterTypeDescriptionDefaultSource
max_tokensintMaximum tokens to generate2048Base
temperaturefloatSampling temperature (0.0-2.0)1.0Base
top_pfloatNucleus sampling threshold0.95Base
seedintRandom seed for deterministic outputsNoneBase
stop_sequenceslist[str]Sequences that stop generationNoneBase
presence_penaltyfloatPenalty for token presence0.0Base
frequency_penaltyfloatPenalty for token frequency0.0Base
google_safety_settingslist[SafetySettingDict]Content safety configurationNoneSpecific
google_thinking_configThinkingConfigDictThinking behavior configurationNoneSpecific
google_labelsdict[str, str]Billing labels (Vertex AI only)NoneSpecific
google_video_resolution'low' | 'medium' | 'high'Video processing resolutionNoneSpecific
google_cached_contentstrName of cached content to useNoneSpecific