THE SUMMARYAI-generated
Key Concepts
Gemini 2.5 Pro, ChatGPT (free version), Grok 3, AI comparison, image generation, language and creative writing, real-time information retrieval, programming, free AI tools, AI performance evaluation.
Image Generation Test
- Objective: To evaluate the image generation capabilities of Gemini 2.5 Pro, ChatGPT (free), and Grok 3 using a detailed prompt.
- Prompt: A detailed description of a house with a swimming pool.
- Gemini 2.5 Pro:
- Fast processing speed (approximately 5 seconds).
- Good adherence to the prompt details.
- Overall, a solid result.
- Grok 3:
- Slower processing speed (approximately 20 seconds).
- High level of detail and sharpness in the generated image.
- Good adherence to the prompt.
- ChatGPT (Free):
- Slowest processing speed (approximately 3 minutes).
- Excellent detail and sharpness, surpassing Gemini and Grok.
- The speaker suggests the slow speed may be intentional to encourage users to upgrade to a paid version.
- Additional Image Editing Tests:
- Removing sunglasses from a photo: Grok 3 produced the most realistic face, while Gemini removed the sunglasses and part of the hair.
- Adding a vase of flowers to a table: All three performed reasonably well, except Grok 3 had an issue with the vase placement.
- Transforming a photo into a Disney-style cartoon: ChatGPT produced the best result, while Gemini added details that didn't fit, and Grok 3 had an issue with adding two pairs of glasses.
- Changing the aspect ratio of an image from vertical to 16:9: Results were self-explanatory and visually presented.
- Image Generation Test Summary:
- Top 1: Gemini 2.5 Pro: Fast processing, good facial image handling, resolution depends on the original image, no image limit. Weakness: Stability issues, may require multiple attempts.
- Top 2: ChatGPT (Free): High detail and resolution, good text handling in images. Weaknesses: Poor facial image handling, slow processing (3 minutes per image), limited to approximately 3 images per day.
- Top 3: Grok 3: Fast image generation, decent resolution. Weaknesses: Unstable, prone to errors, limited image generation.
Language and Creative Writing Test
- Objective: To assess the language and creative writing abilities of the three AI models.
- Test: Writing a short news article using a common prompt.
- ChatGPT (Free):
- Fastest response time.
- Complete information.
- The writing style was considered slightly robotic.
- Gemini 2.5 Pro:
- Satisfactory results.
- Unique and relevant wording.
- Demonstrated superior writing and style capabilities.
- Grok 3:
- Slower response time.
- Excellent output.
- Serious tone, avoiding humor or lightheartedness.
- Additional Language Tests:
- English to Vietnamese translation: Results were presented visually for audience evaluation.
- Transforming regular text into humorous text: Results were presented visually for audience evaluation based on individual humor preferences.
- TikTok content creation: Results were presented visually for audience evaluation.
- YouTube content creation: Results were presented visually for audience evaluation.
- Language and Creative Writing Test Summary:
- Top 1: Gemini 2.5 Pro: Superior performance, rich language, close adherence to prompts, provides content structure and facial expression suggestions for content creation.
- Top 2: Grok 3: Good structure for news articles, but writing style doesn't fully match the prompt and lacks some fluidity. Acceptable for independent use.
- Top 3: ChatGPT (Free): Fastest response, but only adequate. YouTube content was short and lacked detail compared to Gemini and Grok. Suitable for reference, but not for direct use.
Real-Time Information Retrieval Test
- Objective: To evaluate the ability of the AI models to retrieve real-time information.
- Test 1: Asking for the weather in Hanoi on October 1, 2025 (the day of recording).
- ChatGPT (Free):
- Provided a weather forecast starting from 11:00 AM, indicating real-time information retrieval.
- Cited reputable sources, including VTV's YouTube channel.
- Gemini 2.5 Pro:
- Similar to ChatGPT, but sourced information from Google Weather.
- Grok 3:
- Provided basic information, but less detailed than the other two.
- Test 2: Creating a table of gold prices on October 1, 2025.
- Test 3: Finding the latest video from Vat Vao Studio.
- Test 4: Finding the latest video from Vat Vao Studio using "deep research" mode.
- Real-Time Information Retrieval Test Summary:
- Top 1: Gemini 2.5 Pro: Detailed table creation, strong integration with Google and YouTube. However, recent videos (within the last 4 months) were not updated, affecting the accuracy of the latest video search.
- Top 2: ChatGPT (Free): Excellent detailed table creation with citations from reputable sources like VTV. However, thumbnail images for YouTube video searches were not always accurate, and video links were sometimes inconsistent.
- Top 3: Grok 3: Effective when using the "deep search" feature. However, it showed weaknesses in retrieving local Vietnamese information (weather, Vat Vao Studio), possibly due to its data sources being primarily based on X (formerly Twitter).
Programming Test
- Objective: To assess the programming capabilities of the AI models.
- Test 1: Programming a simple airplane game.
- ChatGPT (Free) and Gemini 2.5 Pro:
- Generated code quickly (approximately 10 seconds).
- Grok 3:
- Took longer to process.
- Gemini 2.5 Pro:
- Offered a "preview" feature to immediately test the game.
- Stable and effective.
- ChatGPT (Free):
- Required pasting the code into a web environment for testing.
- The helicopter design was acceptable, but the game experience was unstable.
- Grok 3:
- Similar to ChatGPT, the game interface and experience were not well-developed.
- The airplane couldn't take off.
- Test 2: Programming the "dinosaur jumping" game (commonly seen when there's no internet connection).
- Programming Test Summary:
- Top 1: Gemini 2.5 Pro: Highly effective, intelligent AI with a unique "preview" feature for immediate testing without external tools.
- Top 2: Grok 3: Known for strong programming capabilities (as demonstrated in a previous Flashy Bird game video). Requires more detailed and specific prompts for optimal results.
- Top 3: ChatGPT (Free): Programming capabilities are limited in the free version. Upgrading to ChatGPT 4 is recommended for better programming performance.
Final Thoughts
- Google is generously offering free access to many of its AI versions, including Gemini 2.5 Pro.
- Gemini 2.5 Pro excels in programming, image generation, and writing style, making it a top choice for users seeking overall effectiveness.
- Combining different AI tools like Grok 3 and ChatGPT (free) can optimize daily tasks and provide significant efficiency.
- The speaker encourages viewers to share their experiences with free AI versions and leave a like and subscribe.
Synthesis/Conclusion
The video provides a detailed comparison of three free AI tools: Gemini 2.5 Pro, ChatGPT (free), and Grok 3. It evaluates their performance in image generation, language and creative writing, real-time information retrieval, and programming. Gemini 2.5 Pro emerges as a top performer, particularly in programming and creative tasks, due to its unique features and overall effectiveness. While ChatGPT and Grok 3 have their strengths, they also have limitations in the free versions. The speaker suggests that users can leverage the strengths of each tool to optimize their workflows.
AI summaries can miss context or contain errors. Check important details against the original video.