PGVECTOR IS NOW FASTER THAN PINECONE. And 75% cheaper thanks to a new open-source extension – introducing pgvectorscale. 🐘 What is pgvectorscale? Pgvectorscale is an open-source PostgreSQL extension that builds on pgvector, enabling greater performance and scalability (keep reading for the actual numbers). By using pgvector and pgvectorscale, developers can build more scalable AI applications, benefiting from higher-performance embedding search and cost-efficient storage. 📈 How does it perform? On our benchmark of 50 million Cohere embeddings (768 dimensions each), PostgreSQL with pgvector and pgvectorscale achieves 28x lower p95 latency and 16x higher query throughput compared to Pinecone for approximate nearest neighbor queries at 99 % recall, all at 75 % less cost when self-hosted on AWS EC2. We also tested it against Pinecone’s p2 high performance index, see the blog post at the end of this post for full results (spoiler: It’s just as impressive). 🤔 Why did we build pgvectorscale? Our team at @timescaledb built pgvectorscale to make PostgreSQL a better database for AI and to challenge the notion that PostgreSQL and pgvector are not performant for vector workloads. ⚙️How does it achieve such good performance? Pgvectorscale brings specialized data-structures and algorithms for large-scale vector search and storage to PostgreSQL as an extension, including: (1) StreamingDiskANN –  a high-performance, cost-efficient vector search index for pgvector data inspired by research at Microsoft, and (2) Statistical Binary Quantization (SBQ), developed by Timescale’s own researchers to improve upon standard binary quantization techniques. These innovations help PostgreSQL deliver comparable and often superior performance than specialized vector databases like Pinecone. 👏 Big shoutout to @cevianNY and @sql_johnpruitt, two senior staff engineers at Timescale, who worked on these technical breakthroughs. 🧑‍💻 Sounds exciting! How can I get started? Pgvectorscale is open-source under the PostgreSQL license, and free to use on any PostgreSQL database. You can find installation instructions on the pgvectorscale GitHub repository (see end of post).  It’s also available on any database service in Timescale’s PostgreSQL cloud platform. 📚Learn more [1] Pgvectorscale explainer blog: [2] Pgvectorscale github repo: Share this post with your followers to let them know about pgvectorscale and comment your reactions and questions.

RT @avthars: Proud to announce that pgai and @TimescaleDB are sponsoring The 2025 @aiDotEngineer Summit NYC! AI Engineer Summit the #1 tec…

RT @avthars: I'm becoming more and more convinced that this the ultimate vibe coder / builder's AI stack. Total cost is $245/mo: - v0 fo…

@benhylak @cursor_ai I do this for my server and app. I have one Cusror window open with a Composer Agent for each. But love the idea of them working in tandem on the same codebase vs different ones.

Proud to announce that pgai and @TimescaleDB are sponsoring The 2025 @aiDotEngineer Summit NYC! AI Engineer Summit the #1 technical conference in the world for AI engineers and leadership. The theme of this Summit is Agents at Work, with a special focus on in-production stories. We're super proud to be supporting this event, and I personally couldn't be more excited for the high-quality sessions and hallway discussions that will take place! Are you attending AI Engineer Summit NYC? And building building RAG or Agents with PostgreSQL and Pgvector? If so, come say hi! Myself and the pgai team will be at the Summit and eager to share stories and demo our latest releases (and sneak peaks and upcoming features).

@neondatabase 🤣@sotoish did you write this? Top banter

@benhylak Even v0? Curious: Which tools are you trying to use for UI generation?

We should all take inspiration from o1/ o3 / R1 and schedule time to think in our day. I've been doing this daily for the past 2 weeks and it's amazing the creative breakthrus you can make with focused reflection and giving your subconscious mind time to digest complexity.

“Prompt to app” meets “Webflow / Bubble / WYSIWYG” Very interesting move from Lovable.

RT @avthars: A thread of tips to get 10x value from OpenAI's $200/mo Deep Research

@DevToD4 Very well said. I think the argument for it is that the extra 20% in results makes a difference. And also if you can afford it then the cost isn’t a problem.

@DevToD4 Fair, I’m a huge OSS AI advocate but I don’t find, for example, DeepSeek as reliable as o1 Pro. You’re right that you could find free alternatives to all of those.

@EastlondonDev Nice, thanks for the rec

@M_a_x_S_p_e_e_d 🙌

And for the best thing I've read on Deep Research, look no further than this:

@paulelliotco See the last tweet of the thread

I've pulled out what I consider the most useful tips and put them in this thread for easy reading: