70 top startup and company launches for the week of September 22-30, 2026, ranked by verified X views on Launch Archive.
70 launches
Today we are launching research lab Search millions of ads on Meta / Tiktok Organic / Instagram Save swipe files or migrate from other platforms in one click Sorry Foreplay, Gethookd, TrendTrack Your days are over https://t.co/vzmavIO2tx
Congrats to @SpamRoss, @mduvall_, and @numeral (YC W23) on their $100M Series C! Numeral automates sales tax compliance end to end, from registrations and calculations to filings and payments. They handle US sales tax plus VAT and GST across more than 90 countries, all in one
Today we're launching the Specialized Intelligence Index (SII): one destination for real-work benchmarks across industries, built by the teams that use them every day. Hear from Fireworks co-founder @the_bunny_chen on the importance of specialized benchmarks: https://t.co/pGZk6stDOe
Congrats to @CalebPeffer, @ericciarla, @nickscamara_, and @firecrawl on their $75M Series B! Firecrawl turns the web into clean, structured data for AI agents, with more than 1.5 million people building on their API. They're also launching Alexandria, a knowledge library where
Introducing Wispr-flow like AI Dictation, BUT using local AI Send messages, Speak ideas into your iPhone, convert your speech to text instantly All done locally, so no cloud involved, giving you unlimited words :) I'm a solo developer trying hard to change the landscape of AI https://t.co/xq39QaQyiA https://t.co/YQlrOwI0D3
Introducing the Places Library 📚 A collection of 100 high-fidelity, real-world 3D environments built for embodied AI. Instantly access real-world 3D environments as simulation-ready USDZ files, ready to use in NVIDIA Isaac Sim, Isaac Lab, and compatible OpenUSD simulators. https://t.co/fL6LUwF8ma
Welcome to Atum. That’s what we tell every new partner when they first touch the network. And now we’re saying it to you. Because today, @atumhq, the open payments network, a first of its kind, is live. We picked the date intentionally. Today, September 22, marks the autumnal https://t.co/dOcDsqu85d
Introducing the Open Axis Benchmark, a living benchmark engine for honest evaluation of robot manipulation models, built together with @openroboto. Robotic models evolve faster every month, while most benchmarks stay frozen. Models overfit to fixed task sets, scores stop https://t.co/VIiiTkwgYq
Today, we're launching Delegate: @dlgtio With Delegate, your Claude / Codex / ChatGPT can do things that were previously impossible. It gives your agent access to data it can't reach on its own: TikTok, Instagram, Reddit, YouTube, LinkedIn, B2B contact databases, financial and https://t.co/wtZrOTRrBS
Today, we’re announcing Medusa Search. Full-text and semantic search, built into Medusa and available on all Medusa Cloud plans from today. Read more in the announcement: https://t.co/dYubAhmYmz https://t.co/vO5U2Tpn0o
Introducing Fotor Agent Skills. One Agent. Six ways to create： TVC Films Editable MG Animations UGC Creator Ads AI Talking Presenters Product Ads Product Launch Videos From cinematic storytelling to marketing content, Fotor Agent turns your ideas into finished creative. And https://t.co/gJBzgWAqNk
Today we’re introducing the next Surface Pro 12-inch and Surface Laptop 13-inch with Snapdragon X2 Plus. Catch a vibe with Bolo the DJ and Surface Pro 12-inch. Pre-order now: https://t.co/QGsmq7ms7A https://t.co/F9cTwJDT8b
Today we are releasing DolphinBench: Mapping the Pareto frontier of agent memory. Coding, tool use, and long-context retrieval already have strong public benchmarks. Memory still mostly gets graded as a quiz: ask a question, judge the answer, maybe report precision and recall. https://t.co/OBqs73vG3n
Today we're releasing Moondream Parakeet Redux and Moondream Parakeet Ultra. Two speech-to-text models based on NVIDIA's Parakeet. 25 languages. They run locally, wickedly fast. Redux is for CPUs. Ultra is for GPUs. https://t.co/JlHYtrkSQ5
Introducing Firedrill. The simulation and testing platform for AI agents. Your agents take actions across tools and apps. How do you catch a failure that only shows up five steps into a workflow? With thousands of paths your customers could take, how many have you actually https://t.co/vMxAY3FMZC
Misconfiguring Turnstile by skipping backend validation leaves sites exposed to bots. Introducing Turnstile Spin. It fixes incomplete setups by using your preferred AI coding agent to wire up server-side verification. https://t.co/iFVHNTNSOJ
The easiest way to prepare students for the work they'll do after university is to give them the tools professionals use. We're launching ElevenLabs for Students, providing 3 months of access to ElevenCreative, ElevenAgents, and the ElevenAPI, plus a year of ElevenReader Ultra. https://t.co/ZsAd4o9hcd
Today we’re releasing new Greptile review tiers: Plus and Apex. Higher tiers have more tools and subagents, can run tasks in parallel, better understand the PR, and catch more bugs. - Base: Thorough code review with full codebase context. - Plus: A deeper review for PRs that https://t.co/DtC7z8KUk8
Today, we’re launching ChatGPT-6 Astra, Sol, and Luna for Sales. Paste your website URL. Gojiberry understands your product, finds customers, and launches outreach across every channel. All in seconds. Inside ChatGPT. This feels unreal. https://t.co/OiubdErTxq
We just released our UEFN game, 'Lumber Tycoon Worlds'! Since making the trailer, we've added boss fights. We are currently looking for personal player feedback, not just data from analytics tools. Please check it out and share your thoughts with us! 🏝️ Code: 1987-1432-2641 https://t.co/eIc1wRjqCe
We just launched HeyMilo Red. https://t.co/Bi3dmVjjJG A standalone cheat detection layer that analyzes recordings for AI-assisted answers, coaching, second screens, proxy candidates and synthetic media, then returns timestamped evidence for human review. API-first and built to https://t.co/fofDsp4ltT
ANTS staking is LIVE The AI economy should reward the people who use it and the people who power it. That’s exactly what the Antseed network does. Choose your provider, stake ANTS, and earn rewards. https://t.co/ywqXvT957S
MiMo-V2.6-Pro is available today as an on-demand deployment. → Highest-scoring open-weight model on the AA Intelligence Index (46) → 1.02T total / 42B active MoE → 1M context, multi-modal, with MIT license Spin up an endpoint and tell us what you’re building → https://t.co/JrXgrZ8AIU
Today is a big milestone for @reppo Our Eval API, self serve management dashboard and on demand evals for AI and agentic workflows is now live ⛽️ At @reppo , we are building Scale AI on chain that scales like Stripe. https://t.co/QOgtCPYnvG
No token for MEADGOD17z, and there never will be one. Introducing @MEADGOD17z — our proof-of-concept agent built to go out into the wild and show people what AI17Z can actually do. Except from 11:00–7:00 UTC. That’s his bedtime. Autonomous agents need sleep too, apparently. A https://t.co/0KL5uny2QM
Meet the newest addition to the BOOX family: the BOOX Tile Series. It is a brand-new line of lifestyle companions and novelties, designed to bring BOOX into more moments of everyday life. #Picco is the first product in the series — a tiny 3.97" eReader, about the size of a https://t.co/dCv60R7vZB
Today we're launching the world's first X influencer agent Tell it what you want to promote and it -> finds creators -> reaches out -> manages campaigns -> handles payments -> gets your content live Stop paying agencies $50k-$100k for launches https://t.co/eFCI7h2ypj
GLM-5.3-FLASHX IS NOW LIVE ON https://t.co/fEDs0n3D2S. 📢 Developed by https://t.co/gsQRWng913, GLM-5.3-FlashX is a speed-optimized native multimodal model built for fast interactive coding and Agent workflows, now available through both https://t.co/fEDs0n3D2S API and Web Chat. https://t.co/pEUywmd3Vw https://t.co/u9CGrB1aFK
Introducing Rainmaker's Comeback Series for CFB & Tennis Two brand new AI models, College Comeback and Tennis Comeback: Deep Value. A winning trade can pay +200% to +1,000%. Wins less often than Backspin+, but pays far more when it hits. https://t.co/nWgS0Ij7KW
GLM-5.3-FLASHX HAS NOW LANDED ON https://t.co/fEDs0n3D2S. 📢 Developed by https://t.co/gsQRWng913, GLM-5.3-FlashX is the speed-optimized version of its native multimodal model, built specifically for high-frequency interactive programming and Agent workflows. 1️⃣ OPTIMIZED FOR https://t.co/Bc01bYoT79 https://t.co/YDACfi7GVM