zvec
Ranked 7 of 10 vector databases by recommenders
People recommend it as a fast, lightweight embedded vector database that runs in-process and is simple to use. The most common caveat is that speed depends on local disk performance.
People recommend it as a fast, lightweight embedded vector database that runs in-process and is simple to use. The most common caveat is that speed depends on local disk performance.
9 people gave an opinion
- recommend it
- 8
- criticize it
- 1
- approval share
- 89%
Recommended in 3 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: not recommended
- May 2026: not recommended
- June 2026: recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 8 people recommend itNeeds 10 or more, not met.
- 8 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 3 of 13 monthsNeeds 4 or more, not met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Fast in benchmarks
“Its benchmark numbers are quite strong, and recent versions also support full-text + hybrid retrieval and DiskANN.”
Comment from 26 Jun 2026 in “Bible as RAG Database” “we're seeing double their throughput—so it's far from trivial on that playing field”
Comment from 15 Feb 2026 in “zvec: embedded vector database” “doubling their throughput at 8K QPS is meaningful in that specific setting”
Comment from 15 Feb 2026 in “zvec: embedded vector database” - Comment from 21 Mar 2026 in “Zvec – A lightweight in-process vector database”
- Comment from 16 Feb 2026 in “zvec: embedded vector database”
- Comment from 12 Feb 2026 in “Zvec: SQLite-like simplicity in an embedded vector database (By Alibaba)”
Simple in-process, cheaper than services
“precomputing embeddings and storing them and running on a small vector search like this somewhere is much simpler/cheaper than running external services”
Comment from 15 Feb 2026 in “zvec: embedded vector database” “The in-process model is really nice when you want to layer it in with your own filtering and ranking”
Comment from 12 Feb 2026 in “Zvec: SQLite-like simplicity in an embedded vector database (By Alibaba)”
What people warn about
Each mark is one comment
Speed depends on local disk
“the gotcha is that the reason zvec is fast is because it is primarily constrained by local disk performance”
Comment from 23 Feb 2026 in “zvec: embedded vector database”
Other vector databases people recommend
The four most recommended other vector databases, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Qdrant | 22 | 3 | 88% | 10 | — | Approved |
| Chroma | 17 | 5 | 77% | 7 | — | Approved |
| sqlite-vec | 17 | 3 | 85% | 7 | — | Approved |
| turbopuffer | 13 | 2 | 87% | 8 | — | Approved |
| zvecThis page | 8 | 1 | 89% | 3 | — | — |
What people said
13 of the 14 comments that mention it, newest first. Each links to the original comment.
“zvec's shared libraries are tens of megabytes in size”
“Its benchmark numbers are quite strong, and recent versions also support full-text + hybrid retrieval and DiskANN.”
“the github shows they're smoking everyone else in the benchmarks”
“Zvec currently uses memory-mapped files (mmap) by default, so once the relevant data is warmed up in the page cache, performance should be nearly identical”
“the gotcha is that the reason zvec is fast is because it is primarily constrained by local disk performance”
“feel free to give it a try”
“with zvec's HNSW implementation, we get ~1ms latency on SIFT”
“doubling their throughput at 8K QPS is meaningful in that specific setting”
“we're seeing double their throughput—so it's far from trivial on that playing field”
“we've applied optimizations like prefetching, SIMD, and a novel batch distance computation (similar to a GEMV operation) that alone gives ~20% speedup”
“precomputing embeddings and storing them and running on a small vector search like this somewhere is much simpler/cheaper than running external services”
“this vector DB is seriously fast”
“The in-process model is really nice when you want to layer it in with your own filtering and ranking”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.