What happened: SGLang and Miles Add Day-0 Support for Qwen3.8

SGLang and Miles added Day-0 support for Qwen3.8-2.4T-A95B, a 2.4 trillion parameter model with hybrid attention architecture.

Why it matters: SGLang and Miles Add Day-0 Support for Qwen3.8

Matters for engineers deploying large language models at scale who need serving infrastructure optimized for complex hybrid architectures on launch day.

Watch out: SGLang and Miles Add Day-0 Support for Qwen3.8

The performance numbers are specific to GB300 GPUs with particular parallelism configurations; results may differ significantly on other hardware or setups.

In the news

SGLang and Miles Add Day-0 Support for Qwen3.8

LMSYS · Published August 12, 2026 · 3 min read

In 30 seconds

What happened: SGLang and Miles added Day-0 support for Qwen3.8-2.4T-A95B, a 2.4 trillion parameter model with hybrid attention architecture.
Why it matters: Matters for engineers deploying large language models at scale who need serving infrastructure optimized for complex hybrid architectures on launch day.
Watch out: The performance numbers are specific to GB300 GPUs with particular parallelism configurations; results may differ significantly on other hardware or setups.

Listen to this summary

Read the full article

qwen

The patterns behind this

Infini-Attention Architecture
Hybrid Secret & Cache Management Pattern

Each one covers how the technique works, when it earns its cost, and where it breaks.

The Agent Architect

One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.

Weekly email, one-click unsubscribe. We only use your address to send the briefing.