Gemini 3.7 Flash

Name: Gemini 3.7 Flash
Author: Google

Multimodal

Google

Gemini 3.7 Flash is Google's flagship fast model released in August 2026, ranking first on long-context, MRCR v2 (8-needle), WebDev Arena, and Harvey LAB-AA benchmarks. It combines a 1M token context window with top-tier retrieval, reasoning, and web development performance at $0.75 per million input tokens.

Key Specifications

Parameters

Context

1.0M

Release Date

August 13, 2026

Average Score

59.8%

API Documentation Results Blog

Timeline

Key dates in the model's history

Announcement

August 13, 2026

Last Update

August 27, 2026

Today

September 8, 2026

Technical Specifications

Parameters

Training Tokens

Knowledge Cutoff

March 31, 2026

Family

Capabilities

MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)

$0.75

Output (per 1M tokens)

$3.75

Max Input Tokens

1.0M

Max Output Tokens

65.5K

Supported Features

Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Other Tests

Specialized benchmarks

WebDev Arena

Arena.ai WebDev Arena Elo (blog); model card lists the same Elo under Code Arena. • Self-reported

79.4%

GDPval-AA

GDPval-AA v2 Elo (max_score 3000). • Self-reported

50.8%

MRCR v2 (8-needle)

128k cumulative average • Self-reported

97.0%

Harvey LAB-AA

Harvey LAB-AA; complex legal workflows. • Self-reported

91.0%

Agents' Last Exam

Pass rate • Self-reported

26.3%

Artificial Analysis

Artificial Analysis Intelligence Index. Score 56. • Self-reported

56.0%

AutomationBench

Private set • Self-reported

30.4%

BioMysteryBench

Human difficult • Self-reported

43.5%

CharXiv-R

With tools • Self-reported

88.7%

DeepSWE 1.1

DeepSWE v1.1 • Self-reported

65.3%

FrontierCode 1.1

FrontierCode 1.1 Main • Self-reported

43.6%

GDP.pdf

Expert PDF document comprehension • Self-reported

34.0%

HLE-Verified

HLE-Verified; multidisciplinary expert reasoning. • Self-reported

53.6%

LABBench2

LABBench2; biology real-world research tasks. • Self-reported

82.1%

LVBench

Long video understanding • Self-reported

85.4%

OSWorld 2.0

Agentic computer use • Self-reported

47.9%

Terminal-Bench 2.1

Agentic terminal coding • Self-reported

85.8%

Terminal-Bench 3.0

General agent capabilities • Self-reported

14.9%

License & Metadata

License

proprietary

Announcement Date

August 13, 2026

Last Updated

August 27, 2026

Compare Gemini 3.7 Flash

All comparisons

vs Gemini 3.8 Flash vs Gemini 3.5 Flash-Lite vs Gemini 3.6 Flash vs Gemini 3.8 Flash Cyber vs Claude Fable 5.1 vs Inkling-Small vs GPT-5.6 Terra vs DeepSeek-V4-Pro-0813

Similar Models

All Models

Gemini 3.5 Flash

Google

Released:May 2026

Price:$1.50/1M tokens

Gemini 3.5 Flash-Lite

Google

Released:Jul 2026

Price:$0.30/1M tokens

Gemini 2.5 Flash-Lite

Google

Best score:0.6 (GPQA)

Released:Jun 2025

Price:$0.10/1M tokens

Gemini 2.0 Flash Thinking

Google

Best score:0.7 (GPQA)

Released:Jan 2025

Gemini 3.6 Flash

Google

Released:Jul 2026

Price:$1.50/1M tokens

Gemini 2.0 Flash

Google

Best score:0.6 (GPQA)

Released:Dec 2024

Price:$0.10/1M tokens

Gemini 3.1 Pro

Google

Best score:0.9 (GPQA)

Released:Feb 2026

Price:$2.50/1M tokens

Gemini 3.8 Flash

Google

Released:Sep 2026

Price:$0.75/1M tokens

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.