Niraj Rajgor
@rajgor_niraj
I build, experiment, and occasionally blog. Follow for Software + AI content. AI Project - https://iwrite.email Portfolio - https://nirajrajgor.com
You might like
Example 2: Prompt Create a hyper-realistic product photo using my exact product image 9:18 (keep all text, logo, and colors 100% intact and fully readable). Make it look like a large amount of [clear water] has been thrown from the right side, creating a wide, powerful [splash]…
![AmirMushich's tweet image. Example 2:
Prompt
Create a hyper-realistic product photo using my exact product image 9:18 (keep all text, logo, and colors 100% intact and fully readable). Make it look like a large amount of [clear water] has been thrown from the right side, creating a wide, powerful [splash]…](https://pbs.twimg.com/media/GzEcM7RWUAAQ2Gx.jpg)
Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.
The speed at which they are shipping is commendable! - Code review - GPT-5-codex model - Resume conversation - Search (default=disable) And all of these is included in $20 plan, definitely worth trying. #Codex #ChatGPT
Excited to release v1 of /review in the Codex CLI! You can now use gpt-5-codex to investigate and find critical bugs in your code. Check it out and let us know if you have any feedback!
Very much looking forward to Claude Sonnet 4.1 Coding with non-reasoning is faster! #Claude #claudecode #Sonnet4
A new model was spotted in Claude Code on reddit Probably Sonnet 4.1 coming soon 👀
NEW AI MODEL ALERT! Grok Code Fast 1, Available on @OpenRouterAI as well. $0.20/M input, $1.50/M output Interestingly, Sonic disappeared! #Grok @grok
Anthropic launched Comet competition ?? In preview mode, only for limited Max plan users for now. #comet #Claude #browsersecurity
We’ve developed Claude for Chrome, where Claude works directly in your browser and takes actions on your behalf. We’re releasing it at first as a research preview to 1,000 users, so we can gather real-world insights on how it’s used.
In my experience using it via website, cursor and codex-cli for last 4 days. GPT-5 thinking : Better at problem solving when provided with enough context. Not on par with Sonnet in UI stuff. Claude-sonnet-4: Still good at UI stuff. #ChatGPT5
GPT-5 requires a different way of prompting. It's much more susceptible to instruction, especially style and tone, and it does better when provided with reasoning, validation, and planning sections. I wrote a guide based on my few weeks of usage on how you should prompt it. 👇
Sama hinting GPT-5 pro queries for GPT5-Plus users, this will interesting to watch out for! #openaigpt5
we are considering giving a (very) small number of GPT-5 pro queries each month to plus subscribers so they can try it out! i like it too. but yeah if you wanna pay us $1k a month for 2x the input tokens feels like we should find a way to make that happen...
Fascinating. Appears this is NOT an acquisition. Windsurf will continue to operate under new leadership
Big welcome to @_mohansolo and others from the Windsurf team joining Deepmind : )
Big welcome to @_mohansolo and others from the Windsurf team joining Deepmind : )
We are pivoting away from doing enterprise AI transformations ("AI-native Palantir"). For now at least. I've shared our key learnings below. I'll share a detailed blog post soon. What's next? We are going to start moving insanely quickly on several other ideas. Stay tuned 😎
Big news! 🚀 @cursor_ai is now available on both Mobile and Web! Mobile was expected, but the web interface? A total surprise and it’s separate from the IDE. So cool! Oh, and there’s a new $60 Pro+ plan (not on the site yet, but available to everyone). Details on blog:…
Exciting read. AI identifying itself as Human 😂 Anthropic always has this cool ideas to experiment. #Claude #claudeai #aiexperiment
New Anthropic Research: Project Vend. We had Claude run a small shop in our office lunchroom. Here’s how it went.
Some more details wrt Deep research API from OpenAI o3-deep-research - Input: $10, Output: $40, cutoff: May, 2024 o4-mini-deep-research - Input: $2, Output: $8, cutoff: May, 2024 Both supports streaming, image/text. Also mentions ability of MCP connector. #OpenAI
Typical google strategy, acquire customer & data by providing free service. 1000 req/day is pretty insane. Keep in mind, they do train on your data. #Gemini
Gemini CLI is our new #OpenSource AI agent that brings the power of Gemini directly into your terminal! Access Gemini 2.5 Pro with 1M token context window, 60 requests/min, and 1,000 requests/day—at no cost with a free Gemini Code Assist license → goo.gle/3HW0jL0
My 80 year old friend didn’t think this was AI. This tiny Open Source voice model is so freaking good!!
United States Trends
- 1. #RomanEmpireByBangChan 13.7K posts
- 2. ROMAN EMPIRE OUT NOW 11.9K posts
- 3. Jayden Daniels 25.3K posts
- 4. Good Monday 25.6K posts
- 5. jungkook 595K posts
- 6. #MondayMotivation 25.8K posts
- 7. Dan Quinn 7,105 posts
- 8. #STARDOM 4,890 posts
- 9. Perle Labs 5,223 posts
- 10. #River 7,216 posts
- 11. Seahawks 38.9K posts
- 12. Jake LaRavia 6,245 posts
- 13. Commanders 50.6K posts
- 14. Sam Darnold 15.3K posts
- 15. 60 Minutes 78.4K posts
- 16. Godzilla 45.3K posts
- 17. #RaiseHail 8,869 posts
- 18. Bronny 15.6K posts
- 19. Snopes 3,441 posts
- 20. Jaxson Hayes 3,416 posts
Something went wrong.
Something went wrong.