Directory / LLM API / LongCat-2.5
LongCat-2.5
Meituan's new flagship model for long-horizon agentic tasks: a 1.6-trillion-parameter MoE (about 48B active per token) with native visual understanding and stronger agentic coding, served on the LongCat API Platform with OpenAI- and Anthropic-compatible endpoints and a 1M-token context.
💲 Pay-as-you-go, prepaid balance (limited-time discount): $0.30/M uncached input, $0.006/M cached input, $1.20/M output. Only successful (HTTP 200) requests are billed; token packs also available. Source: longcat.chat (checked Nov 2, 2026).
Pros & cons
👍 Pros
- Very cheap for its size: $0.30/$1.20 per million tokens (limited-time pricing)
- Drop-in OpenAI and Anthropic API formats, 1M context and 128K max output
- Near the top on Terminal-Bench 2.1 (88.4), BrowseComp (91.9) and MCP Atlas (83.1), per Meituan
👎 Cons
- Benchmarks are vendor-reported, many self-run
- Current pricing is a limited-time discount
- API model listing shows text input/output only for now
Latest release
LongCat-2.5 launch
Meituan released LongCat-2.5, a 1.6T-parameter native multimodal MoE trained end-to-end on domestic Chinese accelerators, with large-scale long-horizon RL for codebase development and full-project builds; available in LongCat chat and the LongCat API from day one. Source ↗
Ranked in
Alternatives to LongCat-2.5
Claude Code
Coding agentAnthropic's agentic coding tool — understands your codebase, edits files, runs commands, and helps you ship faster.
Cursor
AI code editorAn AI coding agent for building ambitious software — agents turn ideas into code so you can hand off tasks.
AI SDK (Vercel)
Agent frameworkA unified TypeScript SDK for building AI apps with modern streaming, fallbacks, and multi-model support.
Ratings are preliminary editorial opinions. Spotted something out of date? Tell us. Listed here? Get the badge. Some links are affiliate links; they never affect rankings.