Hermes merges 419 PRs, decouples provider plugins

// 1h agoPRODUCT UPDATE

Hermes merges 419 PRs, decouples provider plugins

The open-source Hermes agent ecosystem completed a massive development push by merging 419 pull requests in a single day. The standout architectural upgrades include shipping external process model providers as standalone plugins—enabling out-of-tree OAuth and custom inference processes to cleanly populate model pickers across the app—alongside desktop and Bot Mode enhancements that allow GPT Live audio and text-to-speech streams to dynamically follow active bot sessions.

// ANALYSIS

Merging over 400 pull requests in 24 hours is a wild display of velocity that borders on automated madness, but decoupling model providers into standalone external plugins is the exact modular shift open-source agent frameworks need to scale without collapsing under dependency bloat.

• Isolating external process model providers into out-of-tree plugins prevents core framework bloat while letting developers integrate proprietary OAuth and custom inference backends seamlessly.

• Harmonizing GPT Live and TTS with Bot Mode and Desktop moves Hermes closer to an ambient multimodal assistant rather than just another terminal prompt.

• A single-day throughput of 419 merges highlights blistering shipping speed, though absorbing that much change in one batch puts serious pressure on regression testing and ecosystem stability.

// TAGS

hermesagentopen-sourcepluginsdevtoolspeechtts

DISCOVERED

1h ago

2026-09-21

PUBLISHED

1h ago

2026-09-21

RELEVANCE

6/ 10

AUTHOR

iamlukethedev

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

MODEL14m ago

StepFun drops Step 5 Preview 600B MoE

StepFun introduced Step 5 Preview, a 600-billion-parameter Mixture-of-Experts (MoE) model that routes through 27 billion active parameters per token. Equipped with a 1-million-token context window and native multimodal vision support, the model is targeted specifically at high-demand domains including software development and financial analytics. StepFun claims that Step 5 Preview matches the frontier performance benchmarks of GLM 5.3 and Kimi K3 while cutting per-task inference costs by roughly 65%, with open model weights scheduled for public release on October 15.

NEWS14m ago

Claude Opus 5.5 leaks with slashed pricing

Anthropic is reportedly conducting early-access testing on an unannounced Claude Opus 5.5 checkpoint under the internal codename "Claude Wafer EAP." Leaked benchmarks indicate GPT-6 Astra-level capabilities alongside significantly reduced API rates of $4 per million input tokens and $20 per million output tokens.

INFRA2h ago

Boat highlights cloud VMs for browser agents

ASCII's boat (boat.dev) provides lightweight, cost-effective Linux virtual machine sandboxes engineered specifically for long-running autonomous AI agents. Featuring pre-configured environments with Docker, Chrome, a 60fps virtual desktop, and rapid snapshot forking, the platform is designed to support high-throughput, token-efficient browser automation workflows that minimize execution overhead and resource consumption compared to traditional cloud sandboxes.