Loading video...

Video Failed to Load

Go Home

We set a new ARC-AGI-2 SotA: 85.28% using an Agentica agent (~350 lines) that writes and runs code.

agenticasdk's profile picture

Agentica

2,643 subscribers

183,208 views • 6 months ago •via X (Twitter)

Science & Technology
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

In 44 lines of code, we have an actually useful agent that runs entirely locally, powered by Qwen3 30B A3B Instruct
1:22

In 44 lines of code, we have an actually useful agent that runs entirely locally, powered by Qwen3 30B A3B Instruct

Andrew I. Christianson

41,110 views • 1 year ago

New in Durable: Blog Agent. An AI agent that runs your blog for you: - Studies your business - Finds the topics worth writing - Runs competitive and keyword research - Writes posts aligned to your SEO and GEO strategy Set your tone and posting frequency once. It writes the rest on autopilot.
0:39

New in Durable: Blog Agent. An AI agent that runs your blog for you: - Studies your business - Finds the topics worth writing - Runs competitive and keyword research - Writes posts aligned to your SEO and GEO strategy Set your tone and posting frequency once. It writes the rest on autopilot.

Durable

1,201,637 views • 2 months ago

ARC Prize 2025 Winners Interviews Paper Award 3rd Place Isaac Liao shares the story behind CompressARC - an MDL-based, single puzzle-trained neural code golf system that achieves ~20–34% on ARC-AGI-1 and ~4% on ARC-AGI-2 without any pretraining or external data.
37:50

ARC Prize 2025 Winners Interviews Paper Award 3rd Place Isaac Liao shares the story behind CompressARC - an MDL-based, single puzzle-trained neural code golf system that achieves ~20–34% on ARC-AGI-1 and ~4% on ARC-AGI-2 without any pretraining or external data.

ARC Prize

55,608 views • 8 months ago

Watch an AI agent buy a $500 arcade machine. Using just 10 lines of code. With $USDC, via GOAT (Great Onchain Agent Toolkit)
3:59

Watch an AI agent buy a $500 arcade machine. Using just 10 lines of code. With $USDC, via GOAT (Great Onchain Agent Toolkit)

Crossmint

48,530 views • 1 year ago

using tools, Cursor Agent have the capability to reason about when it should - read & write code - search your codebase - call MCP servers - run terminal commands an example where it connects to a postgres database, runs a query and writes to file
0:25

using tools, Cursor Agent have the capability to reason about when it should - read & write code - search your codebase - call MCP servers - run terminal commands an example where it connects to a postgres database, runs a query and writes to file

eric zakariasson

34,917 views • 1 year ago

I built an MCP AI Agent using Gemini Flash 2.5 with access to AirBnB and Google Maps in just 30 lines of Python Code. 100% Opensource Code.
0:40

I built an MCP AI Agent using Gemini Flash 2.5 with access to AirBnB and Google Maps in just 30 lines of Python Code. 100% Opensource Code.

Shubham Saboo

100,349 views • 1 year ago

Introducing Kombai, the first AI agent built for frontend development. Kombai vastly outperforms SOTA models + generic agents in real-world benchmarks. Watch it add a complex new feature to an OSS codebase with 500K+ lines of code. More use cases and links in comments
3:08

Introducing Kombai, the first AI agent built for frontend development. Kombai vastly outperforms SOTA models + generic agents in real-world benchmarks. Watch it add a complex new feature to an OSS codebase with 500K+ lines of code. More use cases and links in comments

Dipanjan Dey

2,075,015 views • 1 year ago

Here is a demo of the most comprehensive research agent for your code docs, powered by Nozomio Labs. It turns your indexed repos and docs into a dynamic tree that the agent can access on the fly, using SOTA tools to reduce code hallucinations. Now using the latest Anthropic model, Claude Opus 4.5.
0:35

Here is a demo of the most comprehensive research agent for your code docs, powered by Nozomio Labs. It turns your indexed repos and docs into a dynamic tree that the agent can access on the fly, using SOTA tools to reduce code hallucinations. Now using the latest Anthropic model, Claude Opus 4.5.

Arlan

28,405 views • 8 months ago

I built a multimodal AI Coding Agent team with multi-agents. It has 3 AI agents working together as a team to generate and execute the code: 1. Coding Agent using o-3 mini 2. Vision Agent using Gemini 3. Code Execution Agent using o-3 mini and E2B 100% Opensource Code.
1:36

I built a multimodal AI Coding Agent team with multi-agents. It has 3 AI agents working together as a team to generate and execute the code: 1. Coding Agent using o-3 mini 2. Vision Agent using Gemini 3. Code Execution Agent using o-3 mini and E2B 100% Opensource Code.

Shubham Saboo

42,277 views • 1 year ago

PBLClaw has crossed Level 1. 8-yo Lakshveer built an autonomous agent for ESP32 that writes and uploads code, then runs the device - controlled entirely through natural language via Telegram. You describe what you want. It builds. It flashes. It runs. Any device. On the fly.
1:29

PBLClaw has crossed Level 1. 8-yo Lakshveer built an autonomous agent for ESP32 that writes and uploads code, then runs the device - controlled entirely through natural language via Telegram. You describe what you want. It builds. It flashes. It runs. Any device. On the fly.

Capt Venkat 🇮🇳

392,052 views • 5 months ago

AI agent does research Wow, Zahid Khawaja just sent me this. He runs an AI agent that conducts product research and writes a summary on the best headphones This is wild, powered by GPT-4
0:41

AI agent does research Wow, Zahid Khawaja just sent me this. He runs an AI agent that conducts product research and writes a summary on the best headphones This is wild, powered by GPT-4

Linus ✦ Ekenstam

530,021 views • 3 years ago

What if we could have *trustworthy* agents that don't just write code, but also do research, understand multimodal content, and perform many practically useful tasks? Today at OpenHands, we released a new agent that gets SOTA or competitive performance on 8 diverse tasks.
4:40

What if we could have *trustworthy* agents that don't just write code, but also do research, understand multimodal content, and perform many practically useful tasks? Today at OpenHands, we released a new agent that gets SOTA or competitive performance on 8 diverse tasks.

OpenHands

19,052 views • 1 year ago

We’ve been working with OpenAI for the past few weeks to test their latest Computer-using Agent model. On our evals, CUA has set a new SOTA. Once integrated with an agent, it can complete long-horizon tasks previously impossible. Try CUA on our playground and Act SDK for free!
2:28

We’ve been working with OpenAI for the past few weeks to test their latest Computer-using Agent model. On our evals, CUA has set a new SOTA. Once integrated with an agent, it can complete long-horizon tasks previously impossible. Try CUA on our playground and Act SDK for free!

Scrapybara

89,674 views • 1 year ago

Introducing CSS Studio. A design tool that lives in your site. Runs in your browser. Writes code with your existing AI agent. All completely local - so no subscription. One price, lifetime updates.
0:44

Introducing CSS Studio. A design tool that lives in your site. Runs in your browser. Writes code with your existing AI agent. All completely local - so no subscription. One price, lifetime updates.

Motion

299,874 views • 4 months ago

I made an agent using GitHub Copilot SDK to help me become a running influencer🏃🏾‍♀️✨ Here's the Python code you can use to build an agent that uses MCP and Skills to: - draft and send email - design pitch decks - make a website Copilot CLI then runs all of this in a workflow🐍
9:41

I made an agent using GitHub Copilot SDK to help me become a running influencer🏃🏾‍♀️✨ Here's the Python code you can use to build an agent that uses MCP and Skills to: - draft and send email - design pitch decks - make a website Copilot CLI then runs all of this in a workflow🐍

Marlene Mhangami

34,502 views • 6 months ago

Claude Code tip: create an agent that specialises in code review and security. Then add a new memory entry with a nickname for it so you can call it anytime. My code review and security agent is called Big Tony.
0:25

Claude Code tip: create an agent that specialises in code review and security. Then add a new memory entry with a nickname for it so you can call it anytime. My code review and security agent is called Big Tony.

Richie

135,574 views • 7 months ago

Type a strategy in plain English. The agent writes the code, runs a backtest with a live progress bar, and plots every entry and exit on a TradingView chart. Hyperliquid BTC perps, six months of candles.
0:58

Type a strategy in plain English. The agent writes the code, runs a backtest with a live progress bar, and plots every entry and exit on a TradingView chart. Hyperliquid BTC perps, six months of candles.

QuantPilot

812,284 views • 3 months ago