Skip to main content
Kaino.dev
Discover
Evals
News
Academics
Insights
Kaino.dev

Discover, evaluate, and compare AI tools, models, and agents.

Explore

  • Discover
  • Evaluations
  • News
  • Academics
  • Insights

Community

  • Twitter
  • YouTube
  • Instagram
Privacy PolicyTerms of Service

© 2026 Kaino.dev. All rights reserved.

Version 1.1.0
Discover/MODELS/Yi-Lightning
Yi-Lightning logo

MODELS

Yi-Lightning

by 01.AI

modelllm01-aiyi
Visit WebsiteDocumentationGitHub

Overview

01.AI Yi-series flagship LLM available through the Lingyiwanwu developer platform.

Details

Yi-Lightning is a Yi-series large language model from 01.AI. 01.AI’s official timeline describes the October 2024 Yi-Lightning release as a 100B-parameter MoE model and states that it ranked sixth globally and first in China on the LMSYS blind test. The 01.AI-authored Yi-Lightning Technical Report describes it as the company’s latest flagship LLM, an enhanced MoE model, and discusses architecture, training, evaluation, RAISE safety work, and availability through the Lingyiwanwu developer platform. The official Yi foundation-model catalog names Yi-Lightning as a global SOTA MoE language model, and third-party API catalog metadata describes the Lingyiwanwu Platform API as OpenAI-compatible with Yi-Lightning among available models.

When to Use

Use Yi-Lightning when you want to evaluate 01.AI’s flagship Yi-series MoE LLM through the Lingyiwanwu developer platform. Use it for model comparisons where 01.AI’s reported LMSYS/Chatbot Arena positioning and the Yi-Lightning technical report are relevant evaluation inputs. Consider it for general text-generation reasoning or coding-oriented workflows only after checking current model availability limits and pricing in Lingyiwanwu platform materials.

Getting Started

  1. Review the Lingyiwanwu platform documentation: https://platform.lingyiwanwu.com/docs
  2. Check the official Yi foundation-model catalog for Yi-Lightning availability: https://www.lingyiwanwu.com/yi
  3. Read the Yi-Lightning Technical Report for architecture
  4. training
  5. evaluation
  6. and safety details: https://arxiv.org/abs/2412.01253
  7. Verify current API model identifiers
  8. pricing
  9. limits
  10. and production requirements in official Lingyiwanwu platform materials before deployment.

Key Features

  • •100B-parameter MoE model according to 01.AI’s official October 2024 timeline.
  • •Described in the 01.AI-authored technical report as 01.AI’s latest flagship LLM and an enhanced MoE model.
  • •01.AI states that Yi-Lightning ranked sixth globally and first in China on the LMSYS blind test at release.
  • •Technical report covers architecture
  • •training process
  • •evaluation
  • •RAISE safety work
  • •and developer-platform availability.
  • •Listed in supplied API metadata as available through the Lingyiwanwu Platform API.

Capabilities

  • •text generation
  • •general reasoning
  • •coding assistance
  • •developer API access
  • •model evaluation

Last updated Jul 20, 2026

Yi-Lightning · Discover · Kaino