Gemini 3 Flash
Google DeepMindDec 17, 2025N/A
Google's speed/value tier — famously outscores Gemini 3 Pro itself on SWE-bench Verified (78.0% vs 76.2%) at roughly a quarter of the input cost. A default recommendation for bulk agentic coding.
- Status
- Generally available
- License
- Proprietary API
- Context window
- 1M
- Max output
- 64K
- Input · $/M
- $0.5
- Output · $/M
- $3
- Cached input · $/M
- $0.13
- Blended · $/M
- $1.13
- Parameters
- Undisclosed
- Inputs
- Text + images, audio, video
Benchmarks
Composite 78.0· mean of published scores
Reasoning & knowledge
GPQA Diamond
no public score
MMLU-Pro
no public score
Humanity's Last Exam
no public score
ARC-AGI-2
no public score
Coding
SWE-bench Verified
78%
dataset best 93.4SWE-bench Pro
no public score
LiveCodeBench
no public score
Agents & computer use
Terminal-Bench 2
no public score
OSWorld 2.0
no public score
BrowseComp
no public score
Math
AIME 2026
no public score
Known strengths
- Beats sibling Pro on SWE-V
- ¼ of Pro input price
- High serving speed