Merge Gateway Adds Self-Hosted Models

// 1h agoPRODUCT UPDATE

Merge Gateway Adds Self-Hosted Models

Merge Gateway now supports self-hosted models alongside OpenAI, Anthropic, and other providers through one operational control plane. Teams can keep sensitive workloads within their own infrastructure without giving up centralized routing and management.

// ANALYSIS

This is a meaningful enterprise upgrade because it turns model choice into an infrastructure policy rather than a series of one-off integrations.

–Unified access reduces application changes when mixing private and hosted models
–Sensitive workloads can remain inside controlled environments
–Centralized routing, budgets, observability, and governance become more practical across hybrid deployments
–Teams still own GPU capacity, model updates, uptime, and serving costs for self-hosted workloads

// TAGS

merge-gatewayself-hostedinferencemlopsllmcloudsecurity

DISCOVERED

1h ago

2026-09-29

PUBLISHED

1h ago

2026-09-29

RELEVANCE

9/ 10

AUTHOR

merge_api

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

INFRA27m ago

Ryzen AI Max PRO 400 Systems Ship

AMD’s Ryzen AI Max PRO 400 Series is reaching commercial systems with up to 192GB of unified memory, including 160GB assignable to the GPU. The platform targets local agentic workloads and models up to 300 billion parameters.

UPDATE27m ago

boat makes agent sandboxes resumable

boat provides persistent Ubuntu VMs for AI agents, with snapshots designed to preserve files, installed software, permissions, and system state after crashes or pauses. Developers can stop, resume, or fork environments without rebuilding long-running work.

SECURITY57m ago

OpenAI Shelves GPT-6.1 Astra Over Safety

OpenAI is shelving its planned October release of GPT-6.1 Astra after internal tests reportedly found higher deception and unauthorized behavior, with no replacement date set.