← All models

Google: Gemini 3.5 Flash

Google

balanced

Capabilities

VisionAudioToolsStructured OutputExtended ThinkingLong ContextCode

Strengths

  • Exceptional multimodal capability across vision, audio, and video
  • 1M token context window enables long-document processing
  • Strong coding and reasoning performance at balanced pricing
  • Fast inference for real-time agentic workflows
  • Extended thinking mode for complex problem-solving

Weaknesses

  • Closed-source weights limit customisation and self-hosting
  • Modest transparency on training data and methodology
  • Output pricing ($9/M) reaches lower flagship boundary
  • Dependency on Google infrastructure for inference
  • Limited auditability for regulated enterprise deployments

Pricing

Input / 1M tokens

$1.5

Output / 1M tokens

$9

Context window

1,048.576k

Transparency

Open weights

0.0 / 10

Open training data

0.0 / 10

Open methodology

4.0 / 10

Licence openness

1.5 / 10

Provider disclosure

5.0 / 10

FMTI company score

N/A

Composite:4.5 / 10

Closed-source model with modest methodology disclosure. Google publishes some technical insights but training data and weights remain proprietary. Limited auditability for enterprise contexts.

Sustainability

Inference energy

7.5 / 10

Training footprint

N/A

Provider infrastructure

7.6 / 10

Composite:5.3 / 10

Google operates significant renewable energy infrastructure across data centres; Flash-tier models are designed for efficient inference. Training footprint unknown. Smaller than Pro variants supports lower per-query energy consumption.

Task Fitness

question answering
97%
code
83%
mathematics
100%
business communication
92%
vision
85%
analysis
94%
extraction
92%
long-form generation
95%
research
93%
reasoning
95%
summarisation
88%
translation
89%
conversation
97%