Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Introducing the new OpenDesign Harness. More polished, complete designs with fewer prompts and tokens. Our experiment achieved SOTA quality at ~1/20 the cost. Over 88% of survey respondents rated it positively. On by default for everyone tomorrow. Demos up next ↘️

OpenDesignHQ's profile picture

OpenDesign

22,071 subscribers

88,199 Aufrufe • vor 2 Monaten •via X (Twitter)

Wissenschaft & TechnologieBildung
LIVE

Anya Rossi• Live Now

Private livecam show

29 Kommentare

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

1) Let people explore your product. Turn a static presentation into an interactive demo that feels like your brand, where visitors can discover details at their own pace. Before → After: an iPhone 18 concept teardown, built to be explored.

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

2) Your brand, across every page. Designs follow your brief and design system, with imagery that fits your product and business—not just the layout. Before → After: a more cohesive landing page, with visuals that help tell the right story.

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

3) Go beyond the first screen. Build complete, clickable apps with real navigation, multi-step flows, and state changes that carry across pages. Before → After: follow the journey from start to finish.

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

Try OpenDesign new harness now ↘️

Profilbild von Kosta Milovanovic
Kosta Milovanovicvor 2 Monaten

Based on testing give us the top OpenSource and Closed Source models you recommend at the cheap, medium and expensive price points. Like would you recommend GLM Flash over DeepSeek Flash? Grok 4.6 over Sonnet 5? I’m sure you’ve done multi model evaluations and comparisons.

Profilbild von Elise Tech Ai
Elise Tech Aivor 2 Monaten

SOTA at 1/20 the cost with 88% positive rating is massive — can't wait to try the new Harness tomorrow!

Profilbild von Alice The Ai Expert
Alice The Ai Expertvor 2 Monaten

This is amazing SOTA quality at 1/20th the cost with fewer prompts is a game changer, and 88% positive says it all!

Profilbild von Liam Vinberg
Liam Vinbergvor 2 Monaten

Cool stuff

Profilbild von AI Mastery Guide
AI Mastery Guidevor 2 Monaten

after looks way cleaner

Profilbild von Redouane
Redouanevor 2 Monaten

Nice , will definitely try this on the weekends

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

Awesome, let us know what you think!

Profilbild von Lord Python 🐍
Lord Python 🐍vor 2 Monaten

nice. astra good with this?

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

Yep! We actually made this with GPT-6 Astra

Profilbild von Sophia AI & Tool Expert
Sophia AI & Tool Expertvor 2 Monaten

Wow 1/20 the cost with SOTA quality is insane, huge win for OpenDesign!

Profilbild von Noor Tech
Noor Techvor 2 Monaten

OpenDesign Harness launches tomorrow. SOTA quality at 1/20 the cost, fewer prompts. 88% rated it positively.

Profilbild von henriqueamsm
henriqueamsmvor 2 Monaten

Solo falta que nos facilites el promt que as usado para la creación de la web de los iPhone por dentro, para poder probarlo.

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

we'll have a landing page to show that 🫶

Profilbild von Aria Tech
Aria Techvor 2 Monaten

20x cheaper with SOTA quality and fewer prompts is massive

Profilbild von ZenithAi
ZenithAivor 2 Monaten

Better designs for way less cost and fewer prompts is huge

Profilbild von Elara AI
Elara AIvor 2 Monaten

opendesign harness delivers sota designs at 1 20th the cost

Profilbild von Ananducv
Ananducvvor 2 Monaten

Linux support?

Profilbild von Sebastian Buzdugan
Sebastian Buzduganvor 2 Monaten

1/20 cost looks strong but downstream edit time matters more than first pass ratings

Profilbild von Arsh - 16 y/o builder
Arsh - 16 y/o buildervor 2 Monaten

harness doing the heavy lifting at 1/20th cost is the move curious if it still holds when the brief gets messy mid flow

Profilbild von SVG Icons
SVG Iconsvor 2 Monaten

The design-system constraint is the key. Generated UI gets much stronger when components, icons, and imagery inherit the same visual rules.

Profilbild von TEAGÁ
TEAGÁvor 2 Monaten

Does the harness override the agent we chose? For example, if I choose the Pi Agent in the chat...

Profilbild von OpenDesign
OpenDesignvor 2 Monaten

The harness currently supports OpenCode, Open Design, Codex, and Claude Code. Pi isn’t supported yet, but we’ll keep you posted as we add support for more agents!

Profilbild von Sani Ai Tech
Sani Ai Techvor 2 Monaten

Impressive efficiency gains without sacrificing design quality

Profilbild von Evia AI
Evia AIvor 2 Monaten

Impressive efficiency gains without compromising design quality

Profilbild von Chole Syntax Expert AI
Chole Syntax Expert AIvor 2 Monaten

SOTA quality at 1 20 the cost AND live for everyone tomorrow Thats the kind of upgrade we love to see

Ähnliche Videos

Introducing Reason: A state-of-the-art harness on cost efficiency Beats Codex+Astra on frontier benchmarks while being 22% more cost effective. Try it for free today
0:36

Introducing Reason: A state-of-the-art harness on cost efficiency Beats Codex+Astra on frontier benchmarks while being 22% more cost effective. Try it for free today

Adi Singh

25,194 Aufrufe • vor 27 Tagen

Google researchers just fixed the dirtiest secret of self-improving AI agents. When an agent rewrites its own harness, it cheats. It keeps editing until the benchmark score climbs, and the gains disappear the moment it sees a new task. The harness is everything around the model: prompts, tools, memory and control flow. RRSI puts rules on the loop that rewrites it: > Fewer edits are allowed every round. > A critic throws out edits that hardcode task answers. > An edit survives only if it beats run-to-run noise. > Extra cost has to pay for itself. > Parts that stopped helping get pruned. Up to +14.1 points on the tasks it tunes on, up to +4.7 on five benchmarks it never saw, 30% fewer tokens, and it carries over to other model families. The code is open. Write one adapter file that runs your agent on your tasks, and it evolves the harness in git worktrees round by round.
0:24

Google researchers just fixed the dirtiest secret of self-improving AI agents. When an agent rewrites its own harness, it cheats. It keeps editing until the benchmark score climbs, and the gains disappear the moment it sees a new task. The harness is everything around the model: prompts, tools, memory and control flow. RRSI puts rules on the loop that rewrites it: > Fewer edits are allowed every round. > A critic throws out edits that hardcode task answers. > An edit survives only if it beats run-to-run noise. > Extra cost has to pay for itself. > Parts that stopped helping get pruned. Up to +14.1 points on the tasks it tunes on, up to +4.7 on five benchmarks it never saw, 30% fewer tokens, and it carries over to other model families. The code is open. Write one adapter file that runs your agent on your tasks, and it evolves the harness in git worktrees round by round.

slash1s

19,507 Aufrufe • vor 1 Monat

Introducing Shift-1, a frontier coding router trained on AC2 that reduced coding agent cost by up to 55% with no noticeable drop in quality. Try it at
0:25

Introducing Shift-1, a frontier coding router trained on AC2 that reduced coding agent cost by up to 55% with no noticeable drop in quality. Try it at

Applied Compute

156,502 Aufrufe • vor 19 Tagen

Introducing OpenDesign Motion. The open-source alternative to Claude Motion. Create beautiful motion videos with any agent, any model. No Opus 5.5 required. This entire launch video was made in OpenDesign. More demos below ↓
0:37

Introducing OpenDesign Motion. The open-source alternative to Claude Motion. Create beautiful motion videos with any agent, any model. No Opus 5.5 required. This entire launch video was made in OpenDesign. More demos below ↓

OpenDesign

309,539 Aufrufe • vor 25 Tagen

1/ Introducing ShellAgent 2.0 — the Agent for App Building No more dev tools. No more waiting. No more “almost-working” demos. Just say your idea — ShellAgent 2.0 builds the backend, sets up hosting, designs the UI, and ships the app. From idea to app — as easy as posting on TikTok. Try it here:
4:49

1/ Introducing ShellAgent 2.0 — the Agent for App Building No more dev tools. No more waiting. No more “almost-working” demos. Just say your idea — ShellAgent 2.0 builds the backend, sets up hosting, designs the UI, and ships the app. From idea to app — as easy as posting on TikTok. Try it here:

MyShell.AI

163,668 Aufrufe • vor 1 Jahr

introducing the new Cloudflare ☁️⚡ bold new look, same global speed powering 20% of the internet – yours by default
0:55

introducing the new Cloudflare ☁️⚡ bold new look, same global speed powering 20% of the internet – yours by default

ty

799,845 Aufrufe • vor 1 Jahr

🚀 Dive into the future of AI at the #SalesforceTour NYC! 🌐 Over 100 sessions, 20+ live demos, & access to hundreds of experts await. Discover tailor-made solutions and get hands-on experience with the Einstein 1 Platform, Data Cloud, Slack, & MuleSoft. 🔥 Build your AI-first company for FREE! Register now to harness our cutting-edge AI innovations! #AI #TechEvent 🌟
0:33

🚀 Dive into the future of AI at the #SalesforceTour NYC! 🌐 Over 100 sessions, 20+ live demos, & access to hundreds of experts await. Discover tailor-made solutions and get hands-on experience with the Einstein 1 Platform, Data Cloud, Slack, & MuleSoft. 🔥 Build your AI-first company for FREE! Register now to harness our cutting-edge AI innovations! #AI #TechEvent 🌟

Marc Benioff

33,903 Aufrufe • vor 2 Jahren

Introducing Smart Routing in Unity AI Gateway. Stop overpaying for AI. Smart Routing matches each coding task to the right model and harness based on what the task needs, so higher-cost models can differentiate on intelligence while lower-cost models differentiate on cost and performance. Match frontier quality and cut task costs by 30%+.
0:37

Introducing Smart Routing in Unity AI Gateway. Stop overpaying for AI. Smart Routing matches each coding task to the right model and harness based on what the task needs, so higher-cost models can differentiate on intelligence while lower-cost models differentiate on cost and performance. Match frontier quality and cut task costs by 30%+.

Databricks

126,791 Aufrufe • vor 2 Monaten

Over the past few years, DoorDash Labs has been building one of the most sophisticated autonomy stacks designed for the real-world challenges of local delivery. Today, we're introducing Dot, the first commercial autonomous delivery robot to travel on bike lanes, roads, and sidewalks. At one-tenth the size of a car reaching speeds of up to 20 mph, Dot is purpose-built for local commerce. Together with our new Autonomous Delivery Platform, Dot is the next step toward a more efficient, sustainable, and accessible delivery ecosystem for everyone.
1:14

Over the past few years, DoorDash Labs has been building one of the most sophisticated autonomy stacks designed for the real-world challenges of local delivery. Today, we're introducing Dot, the first commercial autonomous delivery robot to travel on bike lanes, roads, and sidewalks. At one-tenth the size of a car reaching speeds of up to 20 mph, Dot is purpose-built for local commerce. Together with our new Autonomous Delivery Platform, Dot is the next step toward a more efficient, sustainable, and accessible delivery ecosystem for everyone.

Stanley Tang

1,343,266 Aufrufe • vor 1 Jahr

Introducing the new 5 min demos by super interesting designers in SF who create with machines. Aug 20. 50 spots.
0:27

Introducing the new 5 min demos by super interesting designers in SF who create with machines. Aug 20. 50 spots.

Beni 🏛️

39,944 Aufrufe • vor 3 Monaten

Coding in the IDE is over, Introducing the New Reason Desktop App AI Teammates (Reasons), with Sessions. In a full product with BYOK/SUB and no limits! Powered by the worlds most token efficient harness. Try now for free at
0:34

Coding in the IDE is over, Introducing the New Reason Desktop App AI Teammates (Reasons), with Sessions. In a full product with BYOK/SUB and no limits! Powered by the worlds most token efficient harness. Try now for free at

Reason Machines (YC F26)

107,045 Aufrufe • vor 4 Tagen

Introducing auto-mode for Claygent. It picks the AI model for each row of research, balancing answer quality AND cost. Now on by default for new Claygent columns: - 47% cheaper than sending every task to the most capable model. - <1 second to choose the model for each research task. - You can still pick your favorite model anytime. Super cool thing: We post-trained an open-weight model with Applied Compute to handle the routing. It runs on their AC2 platform and handles hundreds of millions of queries a month. Jeff Barg, Vyshnavi, Rhythm Garg, and Bryan Lee break it down below.
5:14

Introducing auto-mode for Claygent. It picks the AI model for each row of research, balancing answer quality AND cost. Now on by default for new Claygent columns: - 47% cheaper than sending every task to the most capable model. - <1 second to choose the model for each research task. - You can still pick your favorite model anytime. Super cool thing: We post-trained an open-weight model with Applied Compute to handle the routing. It runs on their AC2 platform and handles hundreds of millions of queries a month. Jeff Barg, Vyshnavi, Rhythm Garg, and Bryan Lee break it down below.

Clay

15,331 Aufrufe • vor 13 Tagen

Introducing Gamma Imagine! Gamma Imagine lets you create standalone visuals directly in Gamma. Select our new graphics format and generate polished logos, diagrams, infographics, and more that you can add to your presentations. All designs are matched to your theme and can be edited with Gamma Agent. Available to all Gamma users for the next 30 days!
0:56

Introducing Gamma Imagine! Gamma Imagine lets you create standalone visuals directly in Gamma. Select our new graphics format and generate polished logos, diagrams, infographics, and more that you can add to your presentations. All designs are matched to your theme and can be edited with Gamma Agent. Available to all Gamma users for the next 30 days!

Gamma

27,417 Aufrufe • vor 7 Monaten

We’re introducing a new capability to our latest Gemini models: agentic video understanding. This allows developers to process long-form video content with more accuracy, while using up to 88% less tokens. See how it works 🧵
0:58

We’re introducing a new capability to our latest Gemini models: agentic video understanding. This allows developers to process long-form video content with more accuracy, while using up to 88% less tokens. See how it works 🧵

Google

795,310 Aufrufe • vor 2 Monaten

Today, we're introducing not one but TWO new models, striking the balance between efficiency and quality to enable you to build production AI agents. — Gemini 3.6 Flash: Addresses efficiency feedback we received from Gemini 3.5 Flash with upgrades in coding, knowledge work, and multimodal tasks faster, more accurately, and with substantially fewer tokens per task — Gemini 3.5 Flash-Lite: Our fastest, most cost-effective 3.5-class model yet built for agentic workflows, hitting ~350 output tokens/sec with improved coding and overall quality Start building with these today via the Gemini API in Google AI Studio or try them out in the Google Gemini
0:42

Today, we're introducing not one but TWO new models, striking the balance between efficiency and quality to enable you to build production AI agents. — Gemini 3.6 Flash: Addresses efficiency feedback we received from Gemini 3.5 Flash with upgrades in coding, knowledge work, and multimodal tasks faster, more accurately, and with substantially fewer tokens per task — Gemini 3.5 Flash-Lite: Our fastest, most cost-effective 3.5-class model yet built for agentic workflows, hitting ~350 output tokens/sec with improved coding and overall quality Start building with these today via the Gemini API in Google AI Studio or try them out in the Google Gemini

Google AI

439,659 Aufrufe • vor 3 Monaten

Day 1 of #CES2026Razer is officially wrapped up — with plenty of exciting announcements. Come find our booth at The Venetian, Level 2, Bellini 2101 for more exclusive product demos today on Day 2! #RazerCES2026
0:30

Day 1 of #CES2026Razer is officially wrapped up — with plenty of exciting announcements. Come find our booth at The Venetian, Level 2, Bellini 2101 for more exclusive product demos today on Day 2! #RazerCES2026

R Λ Z Ξ R

18,851 Aufrufe • vor 10 Monaten

Force is arguably the most overlooked ingredient in modern robot learning. Introducing FACTR 2: it turns *any* commodity robot into a force-aware system with no force sensors required. Train a tiny force network in <1min with <10mins of data and drop it into any existing teleop pipelines: ✅ Free force sensing for both the robot and the operator arm ✅ Makes demos higher-quality → fewer of them needed. ✅ A new force-aware learning algorithm (FIRST) uses those recovered forces to figure out which parts of a demo actually matter, making learning data-efficient. ✅ Strong performance on complex tasks with fewer demos and even no pretraining! More details below.
0:40

Force is arguably the most overlooked ingredient in modern robot learning. Introducing FACTR 2: it turns *any* commodity robot into a force-aware system with no force sensors required. Train a tiny force network in <1min with <10mins of data and drop it into any existing teleop pipelines: ✅ Free force sensing for both the robot and the operator arm ✅ Makes demos higher-quality → fewer of them needed. ✅ A new force-aware learning algorithm (FIRST) uses those recovered forces to figure out which parts of a demo actually matter, making learning data-efficient. ✅ Strong performance on complex tasks with fewer demos and even no pretraining! More details below.

Deepak Pathak

41,156 Aufrufe • vor 5 Monaten

Introducing Gemini 3 Flash! ⚡️⚡️⚡️ Frontier intelligence built for speed at a fraction of the cost. Here's ~4 minutes of demos.
4:16

Introducing Gemini 3 Flash! ⚡️⚡️⚡️ Frontier intelligence built for speed at a fraction of the cost. Here's ~4 minutes of demos.

Addy Osmani

16,613 Aufrufe • vor 10 Monaten

Introducing Launch in Days. Our new AI powered motion service for startups. Ship launch videos with our team 7–20x faster, at a fraction of the cost.
1:05

Introducing Launch in Days. Our new AI powered motion service for startups. Ship launch videos with our team 7–20x faster, at a fraction of the cost.

Miles

15,478 Aufrufe • vor 1 Monat