“If AI model companies make the engine, GenSpark is a company that builds the car equipped with that engine to take users to their destination.” (Wen Shang, chief operating officer (COO) of GenSpark)
“When OpenAI builds intelligence, numerous startups create applications that utilize that intelligence. GenSpark is a prime example of this.” (Mark Manara, head of startups at OpenAI)
As competition in the generative artificial intelligence industry rapidly shifts from model performance to AI agents that perform actual tasks, cooperation between global big tech/frontier model companies and AI-native enterprises is entering full swing.
Microsoft (MS), OpenAI, and Anthropic shared their respective collaboration cases with GenSpark during a partner session at the GenSpark Media Tour held in Palo Alto, California, on June 26 (local time), jointly stating, “In the era of AI agents, not only powerful models but also partnerships that connect them to actual work are the core competitiveness.”
The most concrete example of cooperation is MS. In April, MS signed a partnership to integrate GenSpark agents into the Microsoft 365 and Agent 365 environments. It is a structure where users can call AI agents within their existing work environment without navigating to a separate window.
Mayank Wadhwa, president of Microsoft ASEAN, said, “Customers do not want to move back and forth between multiple AI services, but rather want to use AI agents directly within their existing work environment.” Wadhwa added, “When MS's trusted platform is combined with GenSpark's agentic AI, it can provide significantly greater value to customers.”
Anthropic introduced GenSpark as a representative 'AI-native partner.' GenSpark's 'Super Agent' has a structure that coordinates multiple sub-agents, requiring frontier models with higher reasoning capabilities. Anthropic's Claude serves as the brain that determines when and which tools to call, and finds alternatives if an error occurs.
Danny Stein, GTM leader at Anthropic, evaluated, “A company that could not have existed without AI is an AI-native enterprise, and GenSpark is the most representative case.” Stein noted, “Speed, product completeness, and the execution capability to launch into the market and rapidly iterate and improve are the differentiating factors.”
OpenAI explained that GenSpark is a key partner that validates models in an actual service environment. The 'Realtime API' for real-time voice conversation is a representative example. GenSpark utilized OpenAI's multimodal technologies, such as GPT-Image and Realtime API, to implement real-use scenarios like making phone calls, creating presentation materials, and generating images and videos.
Mark Manara, head of startups at OpenAI, explained, “Benchmark performance and a model working well in an actual product are completely different matters.” Manara noted, “In actual services, various factors such as personality, voice, error handling, and long-duration tasks determine product quality, and GenSpark, as an early adopter of the 'Realtime API,' has been co-creating the necessary features.”
Shang, COO of GenSpark, emphasized, “Through a collaborative structure where we test and provide feedback on top-tier models like GPT, Claude, and Gemini even before their release, we have established a system that rapidly reflects new models onto the GenSpark platform the moment they are disclosed to the public.”
Palo Alto =
