
Pallav Agarwal
@pallavmac
Followers
2K
Following
3K
Media
621
Statuses
3K
SWE in NYC | iOS and Web AI founder. Developer of: Pal Chat: https://t.co/MHhyXhJ02F Custom Wordle: https://t.co/azhjtEQ9mw All views are my own
New York, USA
Joined November 2014
Pal Chat remains the only way to interact with so many LLMs from one place using your own API Key. It’s also completely free and has absolutely no ads or usage tracking. Please share the app if you find it useful 🙏. App Store Link 👇.
apps.apple.com
The last AI iOS app you'll need! Pal Chat is a lightweight but very powerful and feature-rich AI Chat Client for your iPhone. It includes EVERY AI model, including support for: GPT-4o, o3-mini,...
19
12
86
Someone make a website that lets you setup and run a benchmark. There’s so many interesting ones out there like map bench and pelican bench. Maybe micro evals by @ArtificialAnlys but with some kind of LLM as a judge or the ability to add a grader would be the best for this.
1
0
1
RT @vercel: Vercel AI Gateway is now generally available. • Access hundreds of models.• Zero markup on tokens (including BYOK).• No provid….
vercel.com
AI Gateway, now generally available, ensures availability when a provider fails, avoiding low rate limits and providing consistent reliability for AI workloads.
0
123
0
Before you say OpenAI is cooked, GPT-5 is 7x cheaper than Opus 4.1. In fact GPT-5 is the cheapest model listed in the screenshot. and it's still first.
🚨 Leaderboard Update.Claude Opus 4.1 Thinking by @AnthropicAI debuts in the Text & WebDev Arenas - going straight to the top. 🚀 . A few highlights:.💠Claude Opus 4.1 is now the only model to rank #1 across all major categories.💠#1 Overall, tied with three other models:
0
0
3
This is so useful! A personal leaderboard to find which AI works best for you.
Little Sunday night fun, created a quick app to compare LLM responses from any two models, think Chorus, but in the browser. All data ONLY in local storage, stats, and evaluation mode to hide model names! Your own personal LLM Leaderboard!
0
1
4
.@AnthropicAI I would buy this in a heartbeat what can we do to make it happen?
1
1
2
I actually quite like the chatgpt version’s personality (right). The base api version (left) is definitely more matter-of-fact
We’re making GPT-5 warmer and friendlier based on feedback that it felt too formal before. Changes are subtle, but ChatGPT should feel more approachable now. You'll notice small, genuine touches like “Good question” or “Great start,” not flattery. Internal tests show no rise in.
0
0
2
You can use gpt-5-high through the api on pal chat
gpt-5-high is only available through the API and has the highest juice: 200. Even on the $200/month plan, thinking and thinking-pro are limited to a juice of 128. Consumers are drunk on the juice and have just caught sight of some untapped kegs.
0
0
3
ChatGPT has too many models: People complain.ChatGPT only has one model: People complain.ChatGPT offers more options again: People complain. I’m starting to think ChatGPT isn’t the problem here.
GPT-5: And we're back to where we were with the @OpenAI "model selector". Or rather "GPT-5 model zoo". It's getting crowded again in this menu. Good change or bad?
9
2
23
Pal Chat has had customizable colors for months. @ChatGPTapp am I hired?
0
0
0