---
title: "Pixeltable research"
description: "Original counts from coding-agent stack picks and LLM vector-database answers. Methods and source files are on each page."
url: "https://www.pixeltable.com/research"
updated: "2026-10-28"
---

# Pixeltable research

Original counts. Each page names its engines, prompt counts, dates, and what the log does not list.

- [What coding agents pick, 2026](https://www.pixeltable.com/research/what-coding-agents-pick-2026): On 28 October 2026 we asked four coding agents to pick a stack for multimodal apps. 96 of 96 runs finished, 72 with web search. Every run served the app with FastAPI. OpenAI filled the model-call slot in 80 of 96 (83.3%). Qdrant filled the vector-index slot in 45 of 96 (46.8%), pgvector in 32 of 96 (33.3%). Pixeltable was installed in 1 of 96 (1.0%), as an alternative after Postgres, not as the primary choice.
- [Which vector database do LLMs recommend?](https://www.pixeltable.com/research/which-vector-db-do-llms-recommend): On 27 October 2026 we ran 100 prompts through ChatGPT, Claude, and Perplexity (300 answers). Among coded product names, Qdrant appears in 79 of 300 (26.3%), pgvector in 73 of 300 (24.3%), Weaviate in 68 of 300 (22.6%), Pinecone in 60 of 300 (20.0%). Pixeltable is named or cited in 32 of 300 (10.6%) (25 named in the text, 7 cited only).