Nova 2Nova 2 release notes
What is new in Nova 2 and the Nova 2 model family.
v2.0.0
Nova 2 Released
Nova 2Multimodal1M contextReasoning
Nova 2 is now available. The same models that understand text, images, video, documents and speech can be used directly in the browser, without installing anything.
What is new
- Multimodal by default — send text, images, video and documents in the same request and get one coherent answer.
- Up to 1 million tokens of context — work with entire codebases, long documents and extended conversations without losing context.
- Extended reasoning — models with reasoning capabilities show their step-by-step analysis for multi-step problems.
- Speech and text — Nova 2 Sonic understands speech input and responds with natural language.
- Google sign-in and credit packs — sign in with Google, then buy credits with a card and start generating.
Models in this release
| Model | Input | Output |
|---|---|---|
| Nova 2 Lite | Text, images, video, documents | Text |
| Nova 2 Sonic | Speech, text | Speech, text |
| Nova Multimodal Embeddings | Text, images, documents, video, audio | Embeddings |
Get started
Open the Nova 2 chat experience or compare credit packs and plans.