# Databricks Lakebase Cache, Together Preemptible Compute, OpenAI Agents API
_Databricks details Lakebase Postgres cache, Together AI halves GPU cost with preemptible compute, and OpenAI launches Agents API for managed cloud agents._
**Published:** 2026-09-13
**Source:** https://www.startuphub.ai/ai-news/artificial-intelligence/2026/briefing-databricks-together-ai-openai-2026-09-13-3cfe
---
Databricks detailed cache improvements for Lakebase Postgres, Together AI cut GPU cluster pricing in half with preemptible capacity and OpenAI launched a managed Agents API. Each targets production costs and operational overhead for teams running data and AI workloads.

## Improving Lakebase Postgres compute cache

Databricks outlined changes to the compute cache for Lakebase Postgres.

The post describes a disaggregated storage model where cache behavior affects flexibility and performance. It points to how the compute layer reuses data to reduce trips to storage without adding operational complexity for teams.

The update matters for Postgres workloads that need scale without manual tuning.

[Databricks](https://www.databricks.com/blog/improving-lakebase-postgres-compute-cache)

## Introducing preemptible compute: the same compute, half the price

Together AI added preemptible compute to GPU Clusters at a flat 50 percent of on-demand pricing. The option carries a five minute drain window and offers the same capacity for interruption tolerant jobs. It gives teams a lower cost path for training and batch inference.

[Together AI](https://www.together.ai/blog/introducing-preemptible-compute-the-same-compute-half-the-price)

## Introducing the Agents API

OpenAI introduced the Agents API as a managed service for building cloud agents. It uses the Codex harness for orchestration, long running sessions and tool use, shifting operational burden off the developer.

[OpenAI News](https://openai.com/index/introducing-the-agents-api)
---
Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.