DiskANN approximate vector indexes are GA in Azure SQL Database, Azure SQL Managed Instance (Always-up-to-date), and Fabric SQL (Hoffman Sep 29; Thota Sep 28). Vectors sit next to operational rows under the same T-SQL/security model—no separate vector store. SQL Server 2025 / MI on SQL Server 2025 policy stay preview.
Databricks GA’d Lakebase Search on AWS/Azure (blog 2026-09-28): lakebase_vector (ANN) + lakebase_text (BM25) in the same serverless Postgres OLTP—hybrid via RRF, no separate search cluster. VectorDBBench/Conexiom figures are vendor-reported.
AWS GA’d direct Iceberg/Parquet lake queries from Aurora PostgreSQL (2026-09-30) via embedded DuckDB and aurora_analytics—no ETL. Join live ops rows (incl. uncommitted) with the lake. PG 17.11+/18.6+; no extra feature fee (compute + S3).
NVIDIA announced Kumo Tabular (HF blog 2026-09-29): open tabular foundation models for cls/reg via structured-data-models. Weights OpenMDW-1.1 on HF; library code Apache-2.0. No Inference Provider—local weights path. The TabArena and #1 claims are NVIDIA’s own.
OpenAI says it disrupted a July 2026 adversarial distillation campaign that sought protected reasoning via ToS-violating interaction patterns—not a crypto/DB breach. Core cluster attributed to individuals associated with Moonshot AI (Kimi); volumes are attempted extractions.
Retrieval quality is decided by the embedding model, the dimensions and the chunking, and it is only known once you measure recall. How to choose, how to test, and where vector search earns its place.
warpfront tagged hipfire v0.4.0 (2026-09-30): Rust LLM inference over HIP kernels for AMD RDNA. Self-reported ≈5,120 tok/s pp8192 on R9700; Qwen3.8-Flash-Next full 262K on one R9700. Serve defaults to 127.0.0.1. Apache-2.0 (residual MIT files).
arXiv 2609.25008v1 is a preprint experience report (not SOTA): after a Candle/Burn failure taxonomy on a ~0.4B Bangla-first run, the author pivoted to train in PyTorch and keep Rust for serving. Self-reported $164 / 54.6h H100; weights not public.
Kling AI says Kling 4.0 Flash is live now for Ultra Yearly subscribers; full Kling 4.0 (up to 4K / 10-bit HDR / ~30s) is pitched for October 2026—not today. Tech Times: Flash is 3–20s at 720p 8-bit SDR. No Kling 4.0 on AA yet.
Utopai Studios launched PAI (production intelligence) and Utopai X on 2026-09-30. On Artificial Analysis AA-Video-T2V v2.0 With Audio (accessed 2026-10-01 ~08:58 IDT), Utopai X (based on MiniMax H3) ranks #2 at Elo 1153 (±11) behind Wan 3.0 at 1159. No API.
AWS S3 Vectors (Sep 30, 2026) adds ENHANCED index mode that filters metadata before similarity search, plus $startsWith. AWS claims up to 5× more matching vectors on highly selective filters—vendor figure, not house-verified. No extra cost; CLASSIC until you upgrade.
Perplexity published open-weight pplx-embed-v2-context-9b-preview on Hugging Face (MIT): contextual chunk embeddings for RAG. Preview may break; not on Perplexity API yet. Use encode_queries vs encode; int8 + MRL 1024/2048; transformers≥5.4 + trust_remote_code.
Google DeepMind announced Gemini 4 Argon on 2026-09-30: phased to Fairwind cyber defenders first, 1M output tokens (not full context), intro $2/$10 then $4/$20 per 1M tokens on the blog—no public model ID or ai.google.dev pricing row yet.
Prompt injection can arrive in a user prompt, web page, file, email or tool result. Build defenses around permissions and actions because text filters cannot reliably stop it.
pgvector, JSONB, full-text search and row security can support an AI app in one PostgreSQL database, with practical limits to measure before you scale.