TAAFT
Free mode
100% free
Freemium
Free Trial
Deals
Companies
Meituan
Models
LongCat Flash Thinking

LongCat Flash Thinking

By Meituan
Model family: LongCat
LongCat-Flash-Thinking is a LongCat-Flash variant optimized as a large reasoning model. It uses a Mixture-of-Experts design where roughly 18.6 to 31.3B of 560B parameters are active per token, balancing efficiency and capability. Trained with a curriculum plus SFT on reasoning and agentic data, it targets difficult tasks such as math, logic, programming, theorem proving and tool-driven workflows.
Text Gen 7
Released: September 23, 2025

Overview

Large reasoning model built on a 560B MoE backbone, activating about 18.6 to 31.3B parameters for advanced chain-of-thought, formal and agentic reasoning tasks.

🔍SEO content💻Coding🔍Research

About Meituan

Chinese local services platform connecting consumers with food delivery, restaurant booking, travel, retail, grocery, and on-demand neighborhood services.

Industry: Internet Marketplace Platforms
Company Size: 111298
Location: Beijing, Chaoyang District, CN
Website: meituan.com
View Company Profile

Other models from this family

View all models from this family

Related Models

  • Gen 7
    BGE small
    BGE small is BAAI’s smallest English BGE v1 embedding model. It is an encoder-only BERT-based text embedder designed to map text into vectors for retrieval, and is positioned as a compact model with competitive performance despite its small size.
    🔍Information retrieval📂Content categorization🔍Conceptual search
    Text
    Released 2y ago
  • Gen 7
    Harrier OSS v1 0.6b
    Harrier OSS v1 0.6B is Microsoft’s multilingual text embedding model for semantic search, retrieval, clustering, similarity, classification, bitext mining, and reranking. It uses a decoder-only architecture with last-token pooling and L2 normalization, supports 94 languages, handles up to 32,768 tokens, and is positioned as a strong multilingual embedding model.
    NewText
    Released 2mo ago
  • Gen 2
    Chinchilla
    Chinchilla is DeepMind’s 70B language model used to demonstrate compute-optimal scaling with more training data and fewer parameters than larger undertrained models.
    Text
    Released 4y ago
Last updated: April 6, 2026
0 AIs selected
Clear selection
#
Name
Task