Sanjay Ramadugu

@cognition @cerebras Am i too late?

Intelligence at 1000 tokens per second, right in your terminal. Now available with SWE-1.6 Fast, powered by @cerebras. We're giving the first 100 people who respond a free month of Max to try it out.

The continuing gap between the capabilities of Gemini Pro 3.1 (very good model) and the capabilities of the Gemini app/website is odd. The model can do what Claude/GPT can do, but there is a minimal harness for tools (file creation, research etc), no auditable CoT/actions, manual canvas, etc. The reason this is odd is that Google is trusted by enterprises & has the compute to burn, so a good harness would solve so many of Gemini’s gaps and make it an easier sell to companies. The model can make Office documents, for example, but the harness doesn’t allow it. It could also decide when to use other Google tools (and Google has a lot of very good AI tools) and apply them, taking advantage of the ecosystem, but it doesn’t consistently. I assume something will be coming out here eventually, but the gap with Claude and ChatGPT has only been growing.

The @MulticaAI Desktop app is good.

The Stitch SDK is already at 1.2k+ stars on GitHub. Can't believe it's only been one week and we already have more ideas cooking. github.com/google-labs-co…

Making the most of Google AI Pro: an honest take on Gemini, Antigravity, and Jules medium.com/p/making-the-m…

Is it just me, or does Claude Code in Claude Desktop feel significantly slower than running it directly in the terminal?

We created the open-source version of Routines in Claude Code. Introducing Multica Autopilot: Run your routines locally with any agent (Opencode, Codex, Hermes, Openclaw). github.com/multica-ai/mul…

@steipete @nakulsinglla Was just playing with acp this last week, interested to get your thoughts on building a framework around it. github.com/sanjay3290/age… @steipete

@nakulsinglla acp subagents are lit

Hey Peter, @steipete you've used Codex a ton, and now you're at OpenAI — any quick tips to really maximize it for coding? Been finding it definitely better than Claude lately (no more handholding/charade BS), but want to squeeze even more out of it. Any tips?🦞

Just shipped a NotebookLM skill that works with any AI agent. It’s a set of robust scripts for querying notebooks and syncing sources. No fragile prompt engineering—just reliable tools your agent can call. Open source: github.com/sanjay3290/ai-…

@_davideast @julesagent Count me in for early testing.

I got something cooking real big for @julesagent. But I need some usage. If it breaks, I'll be right there with a fix. Who wants to be an early tester?

@julesagent Huge unlock.

Repoless comes to the Jules REST API and this has massive potential. With a single POST request you can orchestrate coding tasks to a cloud environment with Node, Python, Rust, Bun, and other runtimes preloaded. We can't wait to see what you all do with this one

Built a Jules skill for any AI coding assistant - Claude Code, Gemini CLI, Cursor, Codex, Goose. Delegate bugs, docs, tests to @julesagent from your terminal. Creates PRs, you pull results. github.com/sanjay3290/ai-… @simpsoka @julesagent

The Claude limits on Antigravity are a joke. You get maybe 20 minutes of real work before you're rate limited. "More generous" doesn't feel generous when you're stuck waiting mid-task. @antigravity @GoogleDeepMind

@liadyosef @AnthropicAI @OpenAI @idosal1 @claudeai @code @goose_oss @ChatGPTapp @liadyosef This is huge! Where's the best place to find docs on submitting apps as connectors to AI clients like Claude or ChatGPT?

MCP Apps are officially here. 🎉 Officially released today by @AnthropicAI, @OpenAI, and MCP-UI (@idosal1 and myself). These apps will run *everywhere*. @claudeai, @code, @goose_oss, even @ChatGPTapp. Everywhere. *Check out the future of the web*: blog.modelcontextprotocol.io/posts/2026-01-…

@kamrify Asked mine to alert me when a stock hits a price. It sent me a Finnhub API link, I gave it the key, it set up a cron and now monitors. You can build this elsewhere, but the ease + using any chat channel as the interface is what sets it apart.

my timeline is full of "clawd bot is INSANE" posts but not a single one sharing an actual real world use case lol

Been running Clawdbot on a Mac Mini before it was cool. After the Mac Mini hype, I moved to a free EC2 instance — full power, zero cost. The best part? I asked Mac Mini Clawd to set up EC2 Clawd. @steipete @openclaw you've built something special.

My AI can tweet now. This is either very smart or very dumb.

After 15 years of lurking, it's my turn 🚀 Used to come here for movie/tech debates. Now? It's my go-to for everything AI. X is becoming the new LinkedIn for AI folks. Time to share my thoughts and make real connections here.