AWS Picks 40 Startups for 2025 Generative AI Accelerator

Amazon Web Services (AWS) has named 40 early-stage companies to its 2025 Generative AI Accelerator (GAIA) — an eight-week hybrid program that pairs startups building foundational generative AI technologies with technical and business mentors, access to AWS tooling, and up to $1 million in AWS credits to scale. 

This year’s selection reflects AWS’s focus on foundational models, agentic workflows, and infrastructure tooling that underpin enterprise AI — areas that are seeing rapid technical progress and strong customer demand. The program is positioned to help startups move beyond pilots to production by improving ML performance, optimising stacks, and sharpening go-to-market playbooks. 

“The pace of generative AI innovation is extraordinary, and it’s happening everywhere,” said Sherry Karamdashti, General Manager & Head of Startups in North America, AWS. “Whether it’s in biotech labs, creative studios, or industrial applications, the pace of generative AI innovation is extraordinary, and it’s happening everywhere.”

APJ and India in the spotlight

AWS selected 10 startups from the Asia-Pacific & Japan (APJ) region for the 2025 cohort, reinforcing APJ as a hotbed for generative AI innovation. “The APJ region has emerged as a powerful hub for generative AI innovation, with the ten selected startups showcasing incredible diversity and technical excellence,” said Tiffany Bloomquist, Head of Startups, Asia Pacific & Japan, AWS.

Three Indian startups made the cut: Hyperbots (finance/F&A agentic AI), Smallest AI (foundational voice models for contact centres), and Stimuler (voice-first English-speaking app for non-native speakers). These firms illustrate the cohort’s practical, enterprise-orientated tilt — from finance automation to voice automation and education-technology use cases. 

Program support and timeline

Selected startups receive a package of support — technical enablement, business mentorship, and up to $1 million in AWS credits — intended to accelerate model training, deployment, and customer proofs-of-value. The 2025 programme launches on October 13 at Amazon HQ1 in Seattle and will wrap up with showcases at AWS re:Invent 2025 in Las Vegas.

What this means for enterprise and B2B founders

For enterprise founders and buyers, the GAIA cohort sends two signals: first, AWS is doubling down on foundational tooling and agentic workflows that enterprises need to operationalise generative AI; second, startups in this cohort are likely to focus on practical, compliance-aware solutions that integrate with existing stacks. That alignment is important for customers evaluating vendor risk and time-to-value. 

Use cases in this cohort — particularly the Indian entrants — skew toward business impact: Hyperbots targets finance process automation with an F&A LLM, Smallest AI aims to upgrade enterprise contact centres with voice models, and Stimuler focuses on scalable, voice-first language training. These are clear examples of generative AI moving from novelty to operational tooling for business workflows.

Risks and what to watch

GAIA provides resources, but winners will be the teams that can:

• Demonstrate model reliability and explainability in regulated settings;

• Show clear integration paths with enterprise data and legacy systems;

• Prove early customer ROI and manage model governance and costs.

Investors and enterprise buyers should watch which startups move from credits and mentorship to paid pilots and scalable contracts over the next 6–12 months.

Click here to see the complete list of 40 startups selected worldwide for the accelerator program.

AWS’s 2025 GAIA cohort amplifies a shift toward enterprise-ready generative AI — with strong APJ representation and three Indian startups that signal practical B2B use cases. The program’s mix of credits, tech access, and mentorship aims to shorten the runway from prototype to production.