Robert Brennan retweetledi
Robert Brennan
219 posts
Robert Brennan
@rbren_dev
CEO of OpenHands https://t.co/rIKNrfewog
Boston Katılım Kasım 2015
292 Takip Edilen701 Takipçiler
Robert Brennan retweetledi
New model release by @nvidia - Nemotron 3 Super!
x.com/ctnzr/status/2…
We got early access to test it in OpenHands and it works well, excited to have a great new locally deployable LLM.
Bryan Catanzaro@ctnzr
Announcing NVIDIA Nemotron 3 Super! 💚120B-12A Hybrid SSM Latent MoE, designed for Blackwell 💚36 on AAIndex v4 💚up to 2.2X faster than GPT-OSS-120B in FP4 💚Open data, open recipe, open weights Models, Tech report, etc. here: research.nvidia.com/labs/nemotron/… And yes, Ultra is coming!
English
Robert Brennan retweetledi
LLMs made generating code cheap. The real bottleneck is verification: checking that the change is actually something you can trust and merge.
To start with this, we trained a critic model that watches your agent work, and verifies the quality of output in real-time.
English
Robert Brennan retweetledi
We built an open-source app that uses AI agents to find and fix security vulnerabilities in your repos!
This helps with:
1. Actually fixing the vulnerabilities (we're already using it ourselves)
2. An example of how to run agents in cloud sandboxes for large-scale refactors
English
@zachlloydtweets Agreed!
It's clear that developers are the first persona to adopt agents at scale--a majority of devs have their own "agent" running on their laptop, though most are using something out-of-the-box
I recently built my own custom OpenClaw-esque dev workstation using OpenHands SDK
English
@kishan_dahya totally agree!
yeah we're able take agents folks have built with the OpenHands SDK and get them running at scale in a Kubernetes cluster
English
i will check it out!
my philosophy is coding agents -> general agents once deployed in a VM with all the skills, integrations, and tools for non technical people.
like when employees would work in a citrix remote desktop
are y’all working with companies to help them deploy these agents like that?
English
Robert Brennan retweetledi
What I do if a paper I like doesn't have a GitHub repo.
-2025: email the authors for the code
2026-: ask openhands to reimplements the code
English
Robert Brennan retweetledi
Three weeks ago we published the OpenHands Index as a comprehensive benchmark of software engineering agents.
Today we have three updates:
- Result visualization with Laminar
- Reward hacking identified and patched w/ Hodoscope
- New language models
x.com/OpenHandsDev/s…
OpenHands@OpenHandsDev
What is the best LLM for agentic software engineering? Today, we're releasing The OpenHands Index to answer this question. It's the first broad-coverage benchmark for AI coding agents, comparing them on accuracy, cost, and runtime across 5 task domains.
English
Robert Brennan retweetledi
Robert Brennan retweetledi
Proud moment for open models.
M2.5 wasn’t designed to be the biggest -
it was designed to scale in real production environments.
Appreciate the deep dive from @OpenHandsDev 🙌
Details: minimax.io/news/minimax-m…
OpenHands@OpenHandsDev
Big news for open models: @MiniMax_AI M2.5 is out and it’s an excellent+affordable coding model. It ranks 4th in our benchmarks, the first open model to beat Claude Sonnet. Only Claude Opus and GPT-5.2 Codex score higher. Details on scores and limited-time free access below 🧵
English
For the next few days, you can use MiniMax 2.5 on OpenHands for FREE!
OpenHands@OpenHandsDev
Big news for open models: @MiniMax_AI M2.5 is out and it’s an excellent+affordable coding model. It ranks 4th in our benchmarks, the first open model to beat Claude Sonnet. Only Claude Opus and GPT-5.2 Codex score higher. Details on scores and limited-time free access below 🧵
English
Robert Brennan retweetledi
Hmm, open models for coding are getting really good. At ability rivaling the closed models and 1/10 the price, it might finally be time to switch.
English
Robert Brennan retweetledi
Can agents break down their own tasks? Not yet.
@rbren_dev, CEO of @OpenHandsDev, explains why certain processes still require human judgment, and where LLMs can actually help.
Watch Brennan's full insights in the latest episode of AI Native Dev.
English
Robert Brennan retweetledi
"We send out agents in parallel to solve CVEs across the entire codebase. 90% of them get resolved automatically."
@rbren_dev, CEO of @OpenHandsDev, on why parallelization beats iterative approaches every time.
Full episode in the comments.
English
Robert Brennan retweetledi
Robert Brennan retweetledi
Learn how to deploy Qwen3-Coder with vLLM and build your own agentic workflows using the @OpenHandsDev SDK, all on your own infrastructure.
🔗 amd.com/en/developer/r…
English
Robert Brennan retweetledi
Making devs finally write docs by calling them “skills” is a great arb
English