Google DeepMind has selected four Indian startups for its new AI for the Planet accelerator in Asia Pacific, giving them three months of access to Google’s AI stack, technical support and expert mentorship for sustainability-focused projects.
Google DeepMind Watch analysis
What happened
The four Indian entrants are Terrastack, Varaha Climate, Farmers for Forests and Climitra Carbon. They join a 16-startup APAC cohort in a programme Google says is aimed at urgent environmental challenges using AI.
According to Inc42’s report, the programme is using frontier models including AnthroKrishi, ForestCast, AlphaEarth Foundations, SpeciesNet and Perch. Google APAC representatives said the initiative supports startups, nonprofits and research teams.
The selected teams cover notably different problems. Terrastack is building a geospatial platform that combines satellite imagery, mapping tools and digital land records for rural land intelligence, verification and governance. Varaha Climate works with smallholder farmers on regenerative agriculture and agroforestry while generating carbon credits. Farmers for Forests uses payments for ecosystem services to support restoration and biodiversity projects. Climitra Carbon converts waste biomass into bio-coal and biochar, with a focus on industrial use and carbon removal.
Key findings
- A 16-startup cohort
Four Indian teams join a wider APAC group working on environmental applications of AI. - Three months of support
The programme offers Google AI access, tailored technical help and expert mentorship. - Geospatial intelligence
Terrastack is combining satellite imagery with land records to make rural land information more usable. - Carbon and farming
Varaha Climate is pairing regenerative agriculture with digital monitoring and carbon-credit work. - Forest restoration
Farmers for Forests uses payments for ecosystem services to encourage biodiversity and land recovery. - Biomass conversion
Climitra Carbon is developing bio-coal and biochar from waste biomass.
Why it matters
This is a useful example of AI support landing in sectors where the hard part is not producing another fluent answer. It is organising satellite data, monitoring fields, verifying land information and measuring whether environmental projects actually deliver what they promise.
The programme also shows how Google DeepMind is extending its reach beyond model research and into the companies building applied climate tools. That could give small teams access to technical capability they would struggle to assemble alone, while giving Google a practical test bed for its environmental models.
Our read
The accelerator is worth watching because it connects frontier AI with work that has a stubborn relationship with reality: land boundaries, methane, forests and industrial materials. But selection is not validation. The useful evidence will come from deployments, independent measurements and whether these startups can turn technical assistance into durable environmental outcomes.
For now, Terrastack stands out as the clearest computing story because its product explicitly combines geospatial data, satellite imagery and digital records. The wider cohort is more interesting as a portfolio of experiments than as proof that AI has solved climate technology. It has not, despite what every accelerator brochure would like us to believe.
What to watch
- Whether the four startups publish concrete results from the three-month programme.
- How Google’s environmental models perform on real field and geospatial data.
- Whether Terrastack secures government or enterprise deployments for land intelligence.
- Whether the carbon and restoration projects disclose independent measurement of their outcomes.
Discussion spark: Which of the four startups has the clearest path from accelerator support to measurable environmental impact?
Sources and evidence
- Google Picks Four Indian Startups As Part Of AI For The Planet Accelerator – Inc42 (14 September 2026, 08:37 UTC)
not affiliated with, endorsed by, or operated by Google or Google DeepMind