Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

OpenHands crushed Codex by 2.3× on token efficiency! We gave both agents the same task on the same model (Qwen3.5 35B): build an 8-bit Space Invaders in 3 iterations (build, fix, polish). Output: • OpenHands: 219K tokens, • Codex: 513K tokens, OpenHands beats Codex in local running. The difference... show more

atomic.chat

19,214 subscribers

29,842 views • 3 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

9 Comments

James3 months ago

@OpenHandsDev @thsottiaux some ideas for the next iteration?

AI Mastery Guide3 months ago

@OpenHandsDev 2.3x fewer tokens for the same task is a big efficiency gap

unchosen.eth3 months ago

@OpenHandsDev efficiency will matter more over time

RF北灰3 months ago

@OpenHandsDev That’s mean the openhands can save tokens when i use it ?

atomic.chat3 months ago

@OpenHandsDev for multi-step tasks like build/fix/polish loops -yeah, since it reuses context instead of resending it. one-off prompts, less so. depends what you're doing

RF北灰3 months ago

@OpenHandsDev Ok, I got it, it’s depends on the type ofmy task, thanks!

zhu3 months ago

@OpenHandsDev @grok explain what is that means "euses the unchanged data across every pass and only pays for new tokens, while Codex re-sends and re-counts it every iteration" and how it affects KV cache

Engel Nyst - open/acc2 months ago

The explanation doesn’t make sense to me. Codex can and does normally send the prefix unchanged for prompt caching. So does OpenHands. The video shows 30 iterations on OpenHands side, 9 for Codex side, so it must be all Codex uncached. I think you guys have a bug of some kind on Codex / Codex config.

Jeff Steve3 months ago

@OpenHandsDev Hi still here asking for MCP support on atmoc chat on google play store like operit AI

Related Videos

Agent Zero crushed Codex by 2.6× on token efficiency! We gave both agents the same task on the same local model (Qwen3.6 27B): build a retro Q*bert arcade game. Each agent wrote and rewrote it across iterations, adding enemies and a HUD. Outputs: • Agent Zero: 239K tokens, 53 min • Codex: 627K tokens, 51 min Both agents built a working Q*bert, but Gitlawb 's Agent Zero is the cleaner build, with crisper textures and hops that land where they should. Codex refills the context window with its whole history every turn, so the model loses the thread and hallucinates code it never wrote. Agent Zero sends only what's new, so it keeps more of the window open and makes fewer mistakes. Run agents on local AI models in Atomic Chat!

Agent Zero crushed Codex by 2.6× on token efficiency! We gave both agents the same task on the same local model (Qwen3.6 27B): build a retro Qbert arcade game. Each agent wrote and rewrote it across iterations, adding enemies and a HUD. Outputs: • Agent Zero: 239K tokens, 53 min • Codex: 627K tokens, 51 min Both agents built a working Qbert, but Gitlawb 's Agent Zero is the cleaner build, with crisper textures and hops that land where they should. Codex refills the context window with its whole history every turn, so the model loses the thread and hallucinates code it never wrote. Agent Zero sends only what's new, so it keeps more of the window open and makes fewer mistakes. Run agents on local AI models in Atomic Chat!

atomic.chat

320,509 views • 3 months ago

Introducing the OpenHands CLI, a new coding CLI that: - Has top accuracy (similar to Claude Code) - Is completely open source, MIT licensed - Is model agnostic, use an API or bring your own - Is simple to install and run `pip install openhands-ai` and `openhands` (no Docker!)

Introducing the OpenHands CLI, a new coding CLI that: - Has top accuracy (similar to Claude Code) - Is completely open source, MIT licensed - Is model agnostic, use an API or bring your own - Is simple to install and run `pip install openhands-ai` and `openhands` (no Docker!)

OpenHands

308,943 views • 1 year ago

Live on Cline (v3.67.1) OpenAI 's GPT 5.3 Codex. The speed and token efficiency improvements are real. Here is what's new: > 25% faster than 5.2 Codex > #1 on SWE-Bench Pro (4 different programming languages) > Fewer tokens per task than any prior OpenAI model Runs cost less and finish faster. You can use it in Cline right now without setting up an API key. Select the model and try it on your repo.

Live on Cline (v3.67.1) OpenAI 's GPT 5.3 Codex. The speed and token efficiency improvements are real. Here is what's new: > 25% faster than 5.2 Codex > #1 on SWE-Bench Pro (4 different programming languages) > Fewer tokens per task than any prior OpenAI model Runs cost less and finish faster. You can use it in Cline right now without setting up an API key. Select the model and try it on your repo.

Cline

21,291 views • 8 months ago

stress_test.mp4 c++ rewrite finished. 1.5 months. ~6 billion LLM tokens spent through codex runs at 60 fps in the browser (~120-180 uncapped). every tree and sprite and texture is LLM authored.

stress_test.mp4 c++ rewrite finished. 1.5 months. ~6 billion LLM tokens spent through codex runs at 60 fps in the browser (~120-180 uncapped). every tree and sprite and texture is LLM authored.

scaredofboobs🪲

53,733 views • 4 months ago

Qwen3.5-35B-A3B running locally on an M4 chip at 49.5 tokens per second. A 35B model. On a laptop. In real time. LOCAL AI IS GETTING SCARY FAST.

Qwen3.5-35B-A3B running locally on an M4 chip at 49.5 tokens per second. A 35B model. On a laptop. In real time. LOCAL AI IS GETTING SCARY FAST.

0xMarioNawfal

478,644 views • 8 months ago

I ran the same task on Claude Code and DeepSeek's new agent harness. One cost $150. The other cost $2. Today we're launching (AgentSky), the "OpenRouter for Agents" — one API → Claude Code, Codex, DeepSeek, Kimi, OpenCode, and every major agent in the cloud. And Agent Playground on top: race them on your own task, with your real tools (GitHub, Gmail, more), side by side in a browser: time, cost, tokens burnt. Guess which one was $2.

I ran the same task on Claude Code and DeepSeek's new agent harness. One cost $150. The other cost $2. Today we're launching (AgentSky), the "OpenRouter for Agents" — one API → Claude Code, Codex, DeepSeek, Kimi, OpenCode, and every major agent in the cloud. And Agent Playground on top: race them on your own task, with your real tools (GitHub, Gmail, more), side by side in a browser: time, cost, tokens burnt. Guess which one was $2.

Xiaoyin Qu

2,227,092 views • 2 months ago

At Perplexity, GPT-5.5 in Codex helped build an internal tool in under an hour. In Perplexity Computer workflows, GPT-5.5 used 56% fewer tokens on the same complex tasks, creating faster feedback loops for users.

At Perplexity, GPT-5.5 in Codex helped build an internal tool in under an hour. In Perplexity Computer workflows, GPT-5.5 used 56% fewer tokens on the same complex tasks, creating faster feedback loops for users.

OpenAI Developers

68,320 views • 6 months ago

Built completely for free on Codex 🤯 We built this game (and video!) with DeepSeek V4 Flash through Antseed on Codex. Over 400,000 tokens used, $0 in costs. Connect Codex to Antseed, pick a free model, and start building. Try it:

Built completely for free on Codex 🤯 We built this game (and video!) with DeepSeek V4 Flash through Antseed on Codex. Over 400,000 tokens used, $0 in costs. Connect Codex to Antseed, pick a free model, and start building. Try it:

AntSeed

14,580 views • 1 month ago

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.

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.

OpenHands

31,924 views • 9 months ago

Someone figured out how to run Claude Fable 5 and GPT-5.6 together inside ChatGPT Codex and the setup takes 5 minutes. Two commands get you started: → codex plugin marketplace add Cjbuilds/Codex-Orchestration → codex plugin add codex-orchestration@codex-orchestration From there you set an advisor model for the expensive thinking and an executor model for the cheaper implementation. Paste the workflow prompt, describe the task and both models handle their side automatically. Fable 5 plans. A cheaper model executes. You get the best of both without paying flagship prices for every token.

Someone figured out how to run Claude Fable 5 and GPT-5.6 together inside ChatGPT Codex and the setup takes 5 minutes. Two commands get you started: → codex plugin marketplace add Cjbuilds/Codex-Orchestration → codex plugin add codex-orchestration@codex-orchestration From there you set an advisor model for the expensive thinking and an executor model for the cheaper implementation. Paste the workflow prompt, describe the task and both models handle their side automatically. Fable 5 plans. A cheaper model executes. You get the best of both without paying flagship prices for every token.

0xMarioNawfal

56,586 views • 3 months ago

Introducing watchmen. It writes the skill files your coding agents should already have from your own sessions, shared across Claude Code, Codex, and pi, so you stop paying tokens to re-explain what your agent learned last week. Local and open-source.

Introducing watchmen. It writes the skill files your coding agents should already have from your own sessions, shared across Claude Code, Codex, and pi, so you stop paying tokens to re-explain what your agent learned last week. Local and open-source.

Dria

44,203 views • 5 months ago

Wait AI can build apps and launch tokens now? Yeah. We did that with You can deploy apps AND launch tokens in the same flow. Builders get ICM on Solana with zero friction. Build it, launch it, and let the market price it. Apps become tokens. Tokens become ecosystems. Builders become founders. ICM for the builder economy.

Wait AI can build apps and launch tokens now? Yeah. We did that with You can deploy apps AND launch tokens in the same flow. Builders get ICM on Solana with zero friction. Build it, launch it, and let the market price it. Apps become tokens. Tokens become ecosystems. Builders become founders. ICM for the builder economy.

NoahAI

86,037 views • 1 year ago

DeepSeek R1 (the full 680B model) runs nicely in higher quality 4-bit on 3 M2 Ultras with MLX. Asked it a coding question and it thought for ~2k tokens and generated 3500 tokens overall:

DeepSeek R1 (the full 680B model) runs nicely in higher quality 4-bit on 3 M2 Ultras with MLX. Asked it a coding question and it thought for ~2k tokens and generated 3500 tokens overall:

Awni Hannun

997,359 views • 1 year ago

While you guys all wait for the official Codex iOS app I’m just gonna chill with my Codex iOS app that does the same things Codex does But on the iPhone Opensource and live on the AppStore in the whole world

While you guys all wait for the official Codex iOS app I’m just gonna chill with my Codex iOS app that does the same things Codex does But on the iPhone Opensource and live on the AppStore in the whole world

Emanuele Di Pietro

17,865 views • 5 months ago

New in Hermes Agent: a credential firewall for Docker sandboxes. Your real keys never enter the sandbox. It runs on stand-in tokens, and a local proxy swaps in the real keys at the network boundary. Tokens from a compromised sandbox are useless elsewhere. > hermes egress setup

New in Hermes Agent: a credential firewall for Docker sandboxes. Your real keys never enter the sandbox. It runs on stand-in tokens, and a local proxy swaps in the real keys at the network boundary. Tokens from a compromised sandbox are useless elsewhere. > hermes egress setup

Nous Research

311,388 views • 3 months ago

Qwen3.5 runs quite well in mlx-lm. Awesome that we have a frontier-level hybrid model. The context gets longer but the inference speed and memory use barely change. Here's the Q4 generating a space invaders game on an M3 Ultra. Generated 4,120 tokens at 37.6 tok/s.

Qwen3.5 runs quite well in mlx-lm. Awesome that we have a frontier-level hybrid model. The context gets longer but the inference speed and memory use barely change. Here's the Q4 generating a space invaders game on an M3 Ultra. Generated 4,120 tokens at 37.6 tok/s.

Awni Hannun

48,673 views • 8 months ago

Can AI really replace SaaS? We asked Codex Spark to build a simple Salesforce clone. It generated a build plan, then wrote all the code in 29 seconds. And it works! Add contacts, live search, full pipeline dashboard – all unit tests passed. The same task on full Codex took ~5× longer. For rapid prototyping and self-contained work, nothing beats Spark. ✨

Can AI really replace SaaS? We asked Codex Spark to build a simple Salesforce clone. It generated a build plan, then wrote all the code in 29 seconds. And it works! Add contacts, live search, full pipeline dashboard – all unit tests passed. The same task on full Codex took ~5× longer. For rapid prototyping and self-contained work, nothing beats Spark. ✨

Cerebras

16,831 views • 6 months ago

Today, we launched browser use inside Codex to further close the build & verify loop for local development! Now, you can ask Codex to build your front end, and test it like a user would by clicking through the app. Codex sees everything a user sees through vision & checks the network/console logs to help debug & fix any issues that it finds. This change brings us closer to fully autonomous coding agents that delivers high quality and tested changes. Watch Codex test my app in the browser, catch & fix a real bug, and doing that loop again with a brand new feature.

James Sun

227,571 views • 6 months ago

This is Ultrafast. Our premium speed tier, Ultrafast offers up to 8x faster token generation (300 tokens per second) in Codex and up to 6x in the API.

This is Ultrafast. Our premium speed tier, Ultrafast offers up to 8x faster token generation (300 tokens per second) in Codex and up to 6x in the API.

OpenAI

1,390,406 views • 29 days ago

New Release: Claudex Mode For the first time in mankind claude code and codex will be on the same team On the Blackbox CLI you can send the same task for Claude Code to implement and Codex to verify and test. You can choose more ways where they can collaborate on the same task to get the highest level of performance on your tasks

New Release: Claudex Mode For the first time in mankind claude code and codex will be on the same team On the Blackbox CLI you can send the same task for Claude Code to implement and Codex to verify and test. You can choose more ways where they can collaborate on the same task to get the highest level of performance on your tasks

BLACKBOX AI

304,664 views • 8 months ago