Gemini 3.1 Pro Preview
google/gemini-3.1-pro-preview
Reasoning-first Gemini preview for agentic coding and complex problem solving
Capabilities
ReasoningYes
Tool callingYes
Structured outputYes
TemperatureSupported
AttachmentYes
Open weightsNo
Limits
Context1,048,576 tokens
Max inputN/A
Max output65,536 tokens
Details
Familygemini-pro
Knowledge cutoff2025-01
Release date2026-02-19
Last updated2026-02-19
Modalities
Inputtext, image, video, audio, pdf
Outputtext
Benchmarks
SWE-Bench Pro
54.2
resolve rate
Terminal-Bench
70.3
success rate
SWE-Bench Pro
46.1
resolve rate
SWE-Atlas Codebase QnA
13.5
score
SWE-Atlas Refactoring
33.81
score
SWE-Atlas Test Writing
29.84
score
Artificial Analysis Coding Agent Index
43
average pass@1
SWE-Atlas Codebase QnA
45.6
pass@1
SWE-Bench Pro
15.1
pass@1
Terminal-Bench
68.3
pass@1
GPQA Diamond
94.3
accuracy
Humanity's Last Exam
44.4
accuracy
ARC-AGI-2
77.1
accuracy
Community Rating
—
No ratings yet
Sign in to rate
Pricing
Cost per 1M tokens
InputN/A
OutputN/A