AhanaAI Company
Pro
$9.99 / month
Vector Database · Durable State · One Runtime
Coming soon. Planned prices below are not a checkout. Email [email protected].
Pricing
Planned monthly prices. Coming soon. Not a checkout.
What it Actually Does
Embeddings and exact lookup beside the session.
- Exact retrieval
- Embeddings beside session context and operational metadata
- A natural fit for RAG memory and control-plane lookup
- One connection, one deployable surface
Keys, counters, queues, and streams that survive a crash.
- Key-value storage with TTL
- Atomic counters for quotas and metering
- Queues and append-only streams
- Compressed durable persistence for recovery and retention
Auth and limits live in this process, not a second box.
- API key authentication with hashed storage
- Rate limiting and payload enforcement
- Self-hosted path from the open repository
- Email us if a paid plan needs a human: [email protected]
Call It · Same Key for State and Vectors · Sample Shape · Not a Live Host
PUT /state/{key} {"ttl": 3600}
PUT /vectors/{key} {"embedding": [0.12, -0.04, 0.88]}
POST /query {"vector": [0.11, -0.03, 0.90], "k": 3}
This runtime: is a vector database with durable state in the same process · Embeddings sit next to keys, queues, and session context · No invented benchmarks on this page.
Use Cases
Agents: tool-call history and embeddings in one runtime, so the next step finds the right note, not a Redis plus a vector box.
Edge: kiosk or POS state that survives a power cut on a small disk, then pulls the matching note back.
One Key · State and Match
State = It Still Remembers · Match = It Finds the Right One
AhanaFlow is a vector database with durable state in the same process · Built for teams that today run Redis plus Pinecone, or Postgres plus pgvector.