EN

Grok Code Fast 1: This new AI Model Bests QWEN 3 Coder and Claude Sonnet 4

#AI Tools
2025-09-04 16:458 min read

Content Introduction

In this video, the speaker discusses Elon Musk's latest AI model named Grock Code Fast One, which is available at a significantly lower cost compared to previous models. The model boasts impressive capabilities and is now set for testing, with an aim to evaluate its performance through various coding tasks using VS Code. The video also demonstrates how to set up and utilize the model, along with showcasing its output in real-time projects, including an autoplay chess game and a Rubik's cube simulator. While Grock performs well on simpler tasks, the video highlights some limitations, particularly with complex projects. The speaker intends to further explore Grock's functionality in future videos, encouraging viewer interaction and feedback on this new AI advancement.

Key Information

  • Elon Musk has released a new AI model called Grock Code Fast One.
  • The model is notably inexpensive, costing $0.20 per million input tokens and $1.50 per million output tokens.
  • Grock Code Fast One has 256,000 contacts, which is deemed sufficient for its functions.
  • In comparison, the previous model, Groc 4, was significantly more expensive at $3 per million input tokens and $15 per million output tokens.
  • The current analysis will compare this model's capabilities with the Quen 3 model, which has proven effective in other tests.
  • Users are encouraged to subscribe to the channel for further updates and tutorial videos.
  • The Grock model can be utilized via the Open Router API, where users can set up their keys to initiate coding tasks.
  • However, the model has limitations, such as not supporting images and browser use.
  • Multiple coding tests will be conducted to evaluate Grock's performance, including creating an autoplay chess game.
  • The Grock model is expected to execute coding tasks, but user intervention may be required for more complex projects.

Timeline Analysis

Content Keywords

Grock AI Model

Elon Musk has launched a new AI model called Grock, which is advertised as being very inexpensive, costing only $0.20 per million input tokens and $1.50 per million output tokens. It has 256,000 contacts and is positioned as a cost-effective alternative to previous models.

Pricing Comparison

Grock's pricing is notably lower than that of its predecessor, Groc 4, which was priced at $3 per million input tokens and $15 per million output tokens. Comparisons are also made to another model, Quen 3, which performs well at nearly half the price.

Grock in Coding Tests

In the video, various coding tests are performed using the Grock AI model within the VS Code environment, showcasing its capabilities and coding performance compared to other AI models.

Performance Testing

The presenter tests Grock by creating a simple autoplay chess program and evaluates its performance during task execution. The model proves to be capable, but some minor issues such as not completing complex tasks are noted.

Limitations

Grock has some limitations, including the inability to support images or browser usage and lacks prompt caching capabilities. It's acknowledged that further improvements may be necessary for complex projects.

Future Plans

The presenter plans to create more videos highlighting Grock’s features, particularly testing it in mobile app development, and invites viewer engagement in the comments about their thoughts on the AI model.

Related questions&answers

What is the name of the new AI model released by Elon Musk?

The new AI model is called Grock Code Fast One.

How much does it cost to use the Grock AI model?

It costs $0.20 per million input tokens and $1.50 per million output tokens.

What are the limitations of the Grock AI model?

The Grock model does not support images or browser use and does not support prompt caching.

How does Grock compare to previous models?

The previous version, Groc 4, cost $3 per million input tokens and $15 per million output tokens.

What kind of projects was Grock tested with?

Grock was tested with a chess project and a Rubik's Cube simulator project.

Was Grock successful in completing the coding tasks?

Grock successfully completed the chess project but faced issues with the Rubik's Cube project.

What does the video suggest about the performance of Grock?

The performance is noted as good for simpler tasks, but it encountered issues with more complex projects.

How can users access the Grock AI model?

Users can access it via the Open Router platform.

What was the cost of API requests during testing?

The overall cost of API requests was $0.04 for the tasks performed.

Will there be more content about the Grock model?

Yes, more videos and projects using Grock are planned.

More video recommendations

Share to: