Skip to main content

Google DeepMindSpecialist modelResearch access

AlphaGenome 2

Summary

AlphaGenome 2 predicts how DNA sequences and variants behave in a specific cell or tissue context. Since November 11, 2026, Google DeepMind has made the system available through an API for non-commercial research. Its documentation specifies sequence windows of up to one million base pairs. This is not a token context window.

Commercial research is in a private preview. AlphaGenome 2 is intended for research and is not approved for clinical or diagnostic applications.

All specifications

Input and output

Input
DNA sequence with cell or tissue context. The documented sequence window supports up to one million base pairs. This is not a token context window.
Source: Input

Watermarking and provenance

AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.

AI watermarking
Unknown
no source
Watermarking technique
Not known
no source

Licence and access

Access
The API is available for non-commercial research. Commercial research is in private preview.
Source: Access
Shutdown
None announced
no source

Further details

Model
AlphaGenome 2. Google DeepMind describes it as the successor to the AlphaGenome system.
Source: Model
Predictions
Regulatory effects of DNA sequences and variants across multiple cellular readouts.
Source: Predictions
Use limitations
Intended for research and not approved for clinical or diagnostic applications.
Source: Use limitations

Prices

API costs
Free for non-commercial use. Query rates vary with demand.Source