Average Ratings 0 Ratings

Total
ease
features
design
support

No User Reviews. Be the first to provide a review:

Write a Review

Average Ratings 0 Ratings

Total
ease
features
design
support

No User Reviews. Be the first to provide a review:

Write a Review

Description

AgentBench serves as a comprehensive evaluation framework tailored to measure the effectiveness and performance of autonomous AI agents. It features a uniform set of benchmarks designed to assess various dimensions of an agent's behavior, including their proficiency in task-solving, decision-making, adaptability, and interactions with simulated environments. By conducting evaluations on tasks spanning multiple domains, AgentBench aids developers in pinpointing both the strengths and limitations in the agents' performance, particularly regarding their planning, reasoning, and capacity to learn from feedback. This framework provides valuable insights into an agent's capability to navigate intricate scenarios that mirror real-world challenges, making it beneficial for both academic research and practical applications. Ultimately, AgentBench plays a crucial role in facilitating the ongoing enhancement of autonomous agents, ensuring they achieve the required standards of reliability and efficiency prior to their deployment in broader contexts. This iterative assessment process not only fosters innovation but also builds trust in the performance of these autonomous systems.

Description

Gemini 4 Argon is a frontier AI model from Google DeepMind built to sustain deep reasoning across complex, long-running professional workflows. Google designed the model for demanding work spanning software engineering, finance, legal tasks, enterprise knowledge work, cybersecurity defense, and creative writing. Argon supports coding, reasoning, multimodality, and multi-step task execution, allowing it to work across workflows that require information gathering, analysis, tool use, and extended problem solving. Its output token limit has been increased from 64,000 to 1 million tokens, giving the model additional capacity for lengthy reasoning and generation within a single trajectory. On DeepSWE v1.1, Google reports a score of 77.9% for real-world long-horizon software engineering, while its AutomationBench score of 51.3% measures performance on end-to-end business workflows. Google also reports strong results on evaluations covering finance, legal work, visual analysis, and long-video understanding, including a 91.7% score on LVBench. For cybersecurity teams, Argon can autonomously discover, validate, and patch software vulnerabilities and achieved a reported 68% score on CWE-bench v1. Google is initially providing the model to selected cyber defenders through its Fairwind Program while strengthening safeguards before expanding access to developers, enterprises, and consumers. Argon is planned to launch at an introductory price of $2 per million input tokens and $10 per million output tokens, with cached input tokens receiving a 95% discount from the standard input price.

API Access

Has API No 

API Access

Has API Yes 

Screenshots View All

Screenshots View All

Integrations

.NET No 
C No 
Cursor No 
Dart No 
Factory Droid No 
Gemini No 
Gemini 3.5 Flash Cyber No 
Gemini 3.8 Flash Cyber No 
Gemini Spark No 
Google No 
Google AI Studio No 
HTML No 
JavaScript No 
JetBrains Junie No 
PHP No 
PowerShell No 
R No 
Replit No 
Swift No 
TypeScript No 

Integrations

.NET Yes 
C Yes 
Cursor Yes 
Dart Yes 
Factory Droid Yes 
Gemini Yes 
Gemini 3.5 Flash Cyber Yes 
Gemini 3.8 Flash Cyber Yes 
Gemini Spark Yes 
Google Yes 
Google AI Studio Yes 
HTML Yes 
JavaScript Yes 
JetBrains Junie Yes 
PHP Yes 
PowerShell Yes 
R Yes 
Replit Yes 
Swift Yes 
TypeScript Yes 

Pricing Details

No price information available.
Free Trial No 
Free Version No 

Pricing Details

$2 per 1M tokens (input)
$2 per million input tokens and $10 per million output tokens, with cached input tokens priced at 95% off input token price.
Free Trial No 
Free Version No 

Deployment

Web-Based Yes 
On-Premises No 
iPhone App No 
iPad App No 
Android App No 
Windows No 
Mac No 
Linux No 
Chromebook No 

Deployment

Web-Based Yes 
On-Premises No 
iPhone App No 
iPad App No 
Android App No 
Windows No 
Mac No 
Linux No 
Chromebook No 

Customer Support

Business Hours Yes 
Live Rep (24/7) No 
Online Support Yes 

Customer Support

Business Hours No 
Live Rep (24/7) No 
Online Support Yes 

Types of Training

Training Docs Yes 
Webinars No 
Live Training (Online) Yes 
In Person Yes 

Types of Training

Training Docs Yes 
Webinars No 
Live Training (Online) No 
In Person No 

Vendor Details

Company Name

AgentBench

Country

China

Website

llmbench.ai/agent

Vendor Details

Company Name

Google

Founded

1998

Country

United States

Website

gemini.com

Product Features

Alternatives

GLM-4.7 Reviews

GLM-4.7

Z.ai

Alternatives