Modal

Modal has been named to the 2026 Enterprise Tech 30 list. The ET30 is an annual list by Wing Venture Capital and Eric Newcomer, voted on by 90+ leading investors and corporate development leaders. It recognizes the private companies with the most potential to shape the future of enterprise technology. Thank you to Wing Venture Capital and Eric Newcomer, and congratulations to all the companies honored this year.

I joined Modal in June of 2025 - the very early innings of our GTM organization. Fast forward 9 months and we just completed our first SKO - Modal Deploy. A few takeaways from the lead up and execution of the event: 1. Revenue-minded Engineers + Product-minded Sellers is an extremely powerful combination Our engineers are incredibly talented presenters and deeply understand our customer's problems. Our GTM team has technical depth and the natural pull to go deeper and learn more. This is a very complementary and powerful combination where 1 + 1 = 3. Akshat Bubna and I are proactively screening for these qualities - we want our respective orgs to feel like one team. _________ 2. We have a massive opportunity ahead of us and it's our job to bend the curve to grow even faster Almost every tailwind is working in our favor - the move to open source models, the proliferation of agents, the shift from AI experimentation to production. But tailwinds don't close deals. We bend the curve by treating GTM like a product: iterating constantly, sharpening our craft, and getting a little better at every interaction. Sales excellence compounds the same way great software does. That's how we increase the odds of winning - one conversation, one deal at a time. _________ 3. Being a generational company starts with the work we do now Modal Deploy didn't come together by accident. It came together because people cared enough to plan for something great. That's the standard. Winning companies don't operate like the company they are today - they operate like the company they're becoming. Every meeting, every deck, every customer interaction is a chance to raise the bar. Rigor isn't something you add at scale. It's how you get there. _________ Proud of every person who made Modal Deploy what it was. Nine months in, I'm more convinced than ever that we're building something special - and we are truly just getting started. Huge shoutout to Lauren Wang, Greta Workman, Akshat Bubna, Erik Bernhardsson, Jonathon Belotti, Rebecka Storm, Charles Frye, Sierra Wallizer, Christopher Davis, Paul Butler, Richard Gong, Sona Dolasia, Paul Licursi, Zach T., Chris Prinz, Monishee Matin, Margaret Shen, Peyton Walters

The team at Runway is pushing the frontier of what's possible in world models. We're proud to work with them to power real-time inference for Runway Characters. Link in comments to read more.

Social engineering attacks evolve constantly. The team at Doppel, an AI-native platform built to detect and disrupt these attacks, needs to evolve faster. Their ML engineers wrote about moving their training and inference workloads to Modal, and what's changed: - Parallelized experimentation - 10x faster image builds - Lower operational overhead

Unstick your AI Modal x Krazam TV

we’re in a very chinese time of our lives… Mindshare of open-source LLM providers has shifted dramatically since a year ago. I estimated proportional share of the most popular LLM model families deployed on Modal by paying users. The numbers corroborate what we’ve heard anecdotally, which is that Chinese open models are dominating the ecosystem. What’s cool is that this data represents users who are self-deploying models, so it's a more accurate snapshot of what's being chosen by businesses & sophisticated AI/ML teams. A few observations of note 🧐 ⭐️ Qwen (Alibaba) has been a major contributor to open models. Their frequency of shipping and coverage of LLMs of all sizes helps explain their ubiquity. Let’s collectively pray that they’ll be business as usual despite their recent leadership shakeup. 📉 Llama (Meta), unsurprisingly, has fallen off a cliff in popularity. TBH I’m surprised they’re represented at all in recent months’ data. 📈 GLM (Z.ai) adoption is rapidly growing. This reflects the surge in activity around AI coding use cases. 🎁 DeepSeek share hasn't expanded, but I expect this to change if DeepSeek v4 is all that it’s hyped up to be. Caveats: these data points are estimates and reflect relative share by number of apps, rather than usage of those apps. They also don't show the long tail of other LLM families being deployed on Modal.

Over 1 billion sandboxes have been launched on Modal. Since launching three years ago, we've seen Modal Sandboxes become foundational to how AI is being built. Today, teams like Lovable, Ramp, Cognition and more are using Modal Sandboxes to power everything from coding platforms and background agents to RL infrastructure at scale.

Deploy Nemotron 3 Super, the latest open model from NVIDIA, today on Modal. Its hybrid architecture can deliver >50% faster token generation compared to the best open models today. We think it'll be used as a higher-throughput alternative to models like gpt-oss-120b. Details: - hybrid Mamba-transformer architecture - 120B parameters with only 12B active - 1M context window Get started with our deployment recipe in the comments below.