LLM Reference

LongCat Models by Meituan LongCat

Meituan LongCatProprietary
1 model2026Up to 1m ctxFrom $0.3/1M input

Details

ResearcherMeituan LongCat
LicenseProprietary
Commercial useCommercial use: conditional
Models1
Released2026
Max context1m

Capabilities

ReasoningAll models
Function CallingAll models
Tool UseAll models

About

LongCat is Meituan LongCat's large language model family. LongCat-2.0 is a large-scale MoE language model with 1.6T total parameters, roughly 48B activated parameters per token, native 1M context support, and public API access through the LongCat API Platform.

Current Variants

Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.

1 in view

Use when the workload needs coding, 1m context, and reasoning.

2026-06coding1m contextreasoning

Release Timeline

1 release group
2026-06
1 current
LongCat-2.0
coding1m contextreasoning
Current

Specifications(1 models)

LongCat model specifications comparison
ModelReleasedContextParametersReasoningFn CallingTool Use
LongCat-2.02026-061m1.6T total; ~48B activeYesYesYes

Available From(1 provider)

Pricing

LongCat model pricing by provider
ModelProviderInput / 1MOutput / 1MType
LongCat-2.0LongCat API Platform$0.3$1.2Serverless

Frequently Asked Questions

What is LongCat used for?
LongCat is used for coding, reasoning, and agent workflows and tool use. The family description and listed model capabilities point to those workloads as the best fit.
How does LongCat compare to Claude 3?
LongCat by Meituan LongCat is strongest where you need coding, while Claude 3 by Anthropic is the closest related family to check for vision and multimodal work. LongCat has 1 listed variant and reaches up to 1m context, while Claude 3 reaches up to 200k context, so compare the specs and pricing tables before choosing a production model.
Which LongCat model should I use?
For the lowest listed input price, start with LongCat-2.0 through LongCat API Platform at $0.3/1M input tokens. For the most capable/latest local choice, evaluate LongCat-2.0 with 1m context and reasoning, tool use, and function calling.