LongCat 2.0 is Meituan's trillion-parameter Mixture-of-Experts (MoE) model built for long-context agentic tasks. It features a 1M-token context window and supports up to 128K output tokens — making it one of the most capable models for document-heavy and multi-step workflows.
The model supports native tool calling, deep reasoning mode (thinking), and prompt caching for cost-efficient high-volume workloads.
API Pricing
Where to Use LongCat 2.0
Long-context document processing
Pass full contracts, research papers, or codebases in a single prompt. The 1M context window eliminates chunking and cross-reference issues.
Agentic workflows and tool calling
With native function calling and multi-step reasoning, LongCat 2.0 can plan, call tools, process results, and iterate without losing track of the original goal.
Complex reasoning tasks
Enable deep-thinking mode for STEM tasks, financial modeling, and logic problems. The model reasons through problems step-by-step before producing a final answer.
High-volume workloads with caching
Prompt caching at $0.0195/1M tokens makes it economical for repeated context patterns — batch pipelines, RAG systems, and multi-turn agents.
Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.
Code understanding, repository-level edits, automated task execution, and agentic workflows are explicitly described as strengths of LongCat-2.0.
It is also documented as suited to whole-repository reasoning and other coding-agent tasks that involve large codebases and cross-file dependencies.
The model is described as a fit for multi-step terminal work, where it can run commands, read errors, and retry until a task passes.
It is additionally presented as useful for research and analysis workflows, including autonomous searching, cross-referencing, structured review generation, and business insight analysis.
import requests r = requests.post( "https://api.aimlapi.com/v1/chat/completions", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "meituan/longcat-2.0", "messages": [ { "role": "user", "content": "Hello!" } ] }, ) print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/chat/completions", { method: "POST", headers: { Authorization: `Bearer ${process.env.AIMLAPI_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ "model": "meituan/longcat-2.0", "messages": [ { "role": "user", "content": "Hello!" } ] }), }); console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/chat/completions \ -H "Authorization: Bearer $AIMLAPI_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"meituan/longcat-2.0","messages":[{"role":"user","content":"Hello!"}]}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
| Cached input |
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| Intelligence | 19.7 | Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Coding | 45.3 | Composite score across standardised coding evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
LongCat 2.0 This page | Reasoning + agents | |||
| Reasoning + agents | ||||
| Balanced coding + agents | ||||
| Long-context, multimodal & agentic workflows | ||||
| Reasoning + agents |
LongCat 2.0 has a 1,000,000 tokens context window and can return up to 128,000 tokens.
LongCat 2.0 takes text as input and returns text.
Use meituan/longcat-2.0 as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
LongCat 2.0 is priced at input $0.975 / 1M tokens, output $3.835 / 1M tokens, cached input $0.0195 / 1M tokens.
Yes, LongCat 2.0 can stream responses as they are generated.
LongCat 2.0 was built by Meituan.
