vllm-kimi-optimization

Analyze Kimi optimization PR diffs in vLLM for runtime compatibility.

721|65|Updated Apr 1, 2026

Search Tags:#model-optimization #vllm #kimi #qa-validation #pr-diff-dossier #runtime-triage

Ready to run

npx skills add https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS --skill vllm-kimi-optimization

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: vllm-kimi-optimization
Source: https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/model-optimization/vllm/vllm-kimi-optimization
Command: npx skills add https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS --skill vllm-kimi-optimization

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

PR-backed optimization guidance for Kimi K2/K2.5/Linear/VL in vLLM to audit, debug, extend, or document Kimi-VL, Kimi-Linear, Kimi-K2.5, Kimi-Audio, parser aliases, and quantized MLA behavior.

Core Features & Use Cases

PR-diff dossiers for Kimi model families to capture diffs and rationale across Kimi-VL, Kimi-Linear, Kimi-K2.5, and Kimi-Audio in vLLM.
Cross-version runtime impact analysis, including TP/DP and MLA quantization considerations, with references to landed PRs and tests.
Use Case: When preparing a production deployment, pull the latest PR diffs from references/pr-history.md to validate compatibility and generate a production readiness dossier.

Quick Start

Load the latest vLLM Kimi optimization dossier and begin by reviewing the PR history in references/pr-history.md, then align local changes with the canonical diffs.

Frequently Asked Questions about vllm-kimi-optimization

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I validate vLLM compatibility for Kimi K2.5 and Kimi-VL models?▼

You can analyze Kimi optimization PR diffs in vLLM to ensure compatibility across Kimi-VL, Kimi-Linear, Kimi-K2.5, and Kimi-Audio. It validates changes against mainline commits and documented production diff dossiers.

What is the best way to analyze tensor parallelism impact for Kimi models in vLLM?▼

Analyzing tensor parallelism impact involves reviewing landed PR diffs that capture runtime path changes. This PR-backed analysis covers tensor parallelism, data parallelism, and MLA quantization considerations for Kimi models.

Does vLLM support quantized MLA behavior for Kimi-Audio and Kimi-Linear?▼

Yes, vLLM supports quantized MLA behavior for Kimi-Audio and Kimi-Linear through landed optimization PRs. You can audit and debug this behavior by aligning local changes with the canonical production diff dossiers.

How do I generate a production readiness dossier for Kimi K2 deployments in vLLM?▼

To generate a production readiness dossier for Kimi K2 deployments in vLLM, pull the latest PR diffs from references/pr-history.md to validate compatibility. This dossier captures diffs and rationale across Kimi model families.

Why do Kimi model parser aliases fail after updating vLLM mainline commits?▼

Kimi model parser aliases may fail after updating vLLM mainline commits due to unaligned local changes. Validate changes against the current mainline commits and the production diff dossiers to resolve this.