# Baseten — Agentic inference optimization: 50-90% faster engines

- Company: Baseten (baseten.co)
- Announced: 2026-10-02T21:09:34+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.baseten.co/blog/agentic-inference-optimization-faster-than-sota
- Record: https://forck.live/items/21696-agentic-inference-optimization-50-90-faster-engines
- Subject: Baseten
- Models affected: Qwen-3.6-35B-A3B

Baseten used the MetaInfer skills-only framework to build a custom inference engine for Qwen-3.6-35B-A3B on a single B200. The engine achieved up to 90% faster single-stream decode speed and 2.33x faster time to first token (TTFT) compared to vLLM 0.25.1, enabling new ultra-low-latency use cases for the model.

## Evidence

Verbatim from https://www.baseten.co/blog/agentic-inference-optimization-faster-than-sota:

> With Qwen-3.6-35B-A3B in NVFP4 precision on a single B200, our LLM-generated inference engine outperformed vLLM 0.25.1 (latest at the time of the experiment) by up to 90% on single-stream decode speed, and with 2.33x faster TTFT, opening up new ultra-low-latency use cases for the model that had not been practical before.

---

Record: https://forck.live/items/21696-agentic-inference-optimization-50-90-faster-engines
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md