# Moss: Sub-10 ms semantic search runtime

> Source: <https://www.moss.dev>
> Published: 2026-07-08 07:26:56+00:00

Preparing your content

Preparing your content

Built for Production AI Systems

Fix it with <10ms search. No vector databases. No latency bottlenecks. Runs directly in browser, edge, device, or cloud.

Used by teams running voice AI, copilots, and real time systems

where milliseconds directly impact user experience.

<10ms

End to end retrieval latency

Up to 100x faster than vector databases

250K+ installs

Used by developers building production AI systems

Across voice, copilots, and real time applications

100% local execution

Offline indexing and querying

No external vector database required

Used in production by teams building real time AI systems

Rethinking retrieval

No external retrieval layer. No network hops. Eliminate latency at the source.

Browser. Edge. Device. Cloud. Deploy where performance matters most.

Enable real time conversational experiences. No lag. No infrastructure overhead.

Developer Experience

Add <10 ms retrieval to your AI stack in a few lines of code

Works with your existing LLM stack including LangChain and Vercel AI SDK.

``` python
from moss import MossClient

client = MossClient(PROJECT_ID, PROJECT_KEY)

docs = [{"text": "How do I track my order?"}]

await client.add_docs("my-index", docs)
```

Benchmarks

Benchmark run on 100K documents. Includes embedding inference and end to end retrieval latency. [View benchmark script](https://github.com/usemoss/moss/tree/main/benchmarks)

Integrations

Drop Moss into your existing stack across voice, LLM frameworks, and frontend AI

Use Cases

For systems where retrieval is on the critical path and latency directly impacts user experience

<10 ms context retrieval for real time conversation. Your agent responds instantly without latency or network overhead.

FAQ

Answers to common questions about latency, architecture, and production deployment