GoKawiilFlide launches open-source inference engine claiming up to 9.5x speedup for Jev decision models
A new open-source project called Flide was shown on Hacker News, offering fast inference infrastructure for Jev-style decision models via an exact state-forking technique. Its creators say it runs up to 9.5 times faster than existing engines vLLM and SGLang while preserving model-only answers, and version 0.1.10 adds support for several new decision models including Ainvest-Decision, Darwin-27B-RSI-Decision and KnowLine-31B-Gen1.5, the latter now able to read images.
GoKawiil's interpretation of the reporting above, not reported fact.
The claimed speed gains could make decision-model deployment cheaper and faster for developers building automated support or classification systems, if the benchmarks hold up outside the project's own tests. Because Flide is positioned as a drop-in alternative to established engines like vLLM and SGLang, its adoption could depend on how well it integrates with existing workflows and how independently verified its performance claims turn out to be. The rapid pace of model additions suggests the project is actively courting a broader ecosystem of decision-model providers.
- Flide claims up to 9.5x faster inference than vLLM and SGLang for Jev-style decision models
- Version 0.1.10 adds Ainvest-Decision, Darwin-27B-RSI-Decision and KnowLine-31B-Gen1.5 with image support
- The project serves models through a /v1/systemone API and was introduced as a Show HN post
Source: github.com, 2026-11-09
Published there as: “Show HN: Flide: Make Jev-like decision models up to 9.5x faster”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.