Tomorrow.io weather forecasting software is considered not worth the cost. Insiders reveal its operation involves data monopolization and suggest focusing on developing local weather stations. They affirm not opposing AI but emphasize the need for responsible use of public funds to ensure value.
Regarding the news on 28 Sep 2026 that the Cabinet approved a 497 million baht budget for the Tomorrow.io project, Chaiyachon Chidchob, Minister of Digital Economy and Society, confirmed today, 29 Sep 2026, that this is fake news. The project is still in the phase of studying and comparing different systems to select the most cost-effective and efficient solution and has not yet been submitted to the Cabinet for approval.
Thailand piloted the Tomorrow.io weather forecasting system for three months from December 2025 to February 2026, with Hat Yai district in Songkhla province as one of the main test sites.
Tomorrow.io is a US-based weather technology company whose business includes not only a weather app but also its own satellites, data from other sources, forecast models, and platforms that convert weather data into alerts or advice for users.
This afternoon, 29 Sep 2026, the Meteorological Department released a summary report stating it partnered with The Tomorrow Companies Inc. to trial a weather forecasting platform using Microwave Sounder satellite technology, supported with free access from December 2025 to 30 June 2026.
The test has now concluded, and the Meteorological Department has ceased using the platform. Evaluation showed the system performed moderately well but was not perfect, requiring further technical improvements and additional functionality.
Thairath Online's special news team spoke with a source working in weather forecasting who revealed that the Tomorrow.io system is not cost-effective and should not be promoted. Instead, investment should focus on serious development of forecasting, possibly integrating AI for more detailed and accurate predictions. However, this carries high risk since AI depends on original data from humans, requiring raw data such as radar and satellite measurements before AI can enhance calculations.
Using Tomorrow.io is like importing end-stage foreign data into Thailand, which is unreasonable. Evaluation of its cost-effectiveness is necessary, especially since Thailand already trialed the system for three months without it meeting operational needs or demonstrating success in forecasting or providing accurate, timely alerts.
"Real data must come from sources like the Meteorological Department, including radar and satellite measurements. Suddenly relying on foreign AI is like using only the end part of their system. If we lack funds next year, it will be abandoned altogether."
Funds should be directed toward developing supercomputers or high-performance processing systems to obtain more detailed variables, then use Thai AI systems for further processing.
Additionally, training and small-scale trials (e.g., 10-20 million baht) should be conducted in specific high-risk areas. If successful, expansion nationwide can follow. The Meteorological Department should focus on improving communication language, public relations, timely alerts, and social media language rather than purchasing expensive forecasting software.
The source said they researched Weathernews Inc., a major Japanese private weather data company providing accurate, specialized B2B weather data and known for its 24-hour Weathernews LIVE broadcast. This system offers flexibility for study and use. If it meets requirements, it can be purchased for just over 1 million baht or by individual function at reasonable prices. If it proves unsatisfactory, the service can be canceled. In contrast, Tomorrow.io's proof of concept tests did not develop prototypes according to the Meteorological Department's needs (e.g., detailed at household or village level) but focused on products the company wanted to produce, making Thailand only a user. If the vendor discontinues service, Thailand will be left without knowledge or systems.
Windy charges about 700 baht per year and offers comprehensive products covering wind and rain.
WeatherNews offers trial use and joint study, with transparent background and service fees just over 1 million baht annually. Users can cancel if dissatisfied.
Tomorrow.io is priced at 497 million baht but provides only an end-user software or application relying on foreign data.
Alternatively, if existing radar and satellite measurement data were processed into localized information with AI simulating scenarios to issue public alerts, it would be more effective.
The source believes officials and the project specification and acceptance committees bear high risks because the system does not meet the Meteorological Department's real needs and may face later scrutiny for cost-effectiveness. Staff must endure confusion from imposed policies, even though personnel and existing systems remain capable. This reflects lessons from the AI Passport system that faced similar issues with purchased AI systems failing to deliver expected practical performance.