There’s a simpler way to ground your AI.
Antfly is the multimodal database with the inference engine built-in. That means all of your data in all of its formats is always searchable.
Free and open source under Apache-2.0
Loved by developers at
- dbt Labs
- Vercel
- HashiCorp
- Datadog
- Tailscale
- GitHub
- GitLab
- Netlify
- NVIDIA
- AWS
- Meta
- Adobe
- Intel
- Qualtrics
- Merck
The multimodal pipeline with zero pipes.
Reading, embedding, indexing, reranking: it all happens inside the engine as data arrives. There’s nothing to wire together and nothing to keep tuned.
# Install and start a local single-node engine, free
curl -fsSL https://releases.antfly.io/antfly/latest/install.sh | sh
antfly standaloneBring whatever data you have.
Start a new project on Antfly, or point it at the Postgres, buckets, and drives you already run. Nothing to migrate, nothing to replace.
PDFs, screenshots, tables, and audio are indexed for you automatically.
Vector, keyword, and graph search are indexes inside one database, fused in a single query plan. The inference that chunks, embeds, and reranks runs in there too. That’s where the pipes went.
Antfly is a single binary. Inference runs inside the engine, so none of your data has to leave your network.
Built-in models
Every modality needs its own models: OCR for scanned pages, transcription for recordings, captions and embeddings for images, an extractor for the graph. Antfly runs them inside the engine, where they build the indexes automatically.
Run it yourself, or let us.
mul·ti·mod·al
your data in all its formats, like text, images, video, and audio.