# LongCat 2.5 now available on AI Gateway

Source: [Vercel](https://vercel.com/changelog/longcat-2-5-now-available-on-ai-gateway)  
Feed7 permalink: https://feed7.dev/p/longcat-2-5-now-available-on-ai-gateway-18qougo  
Published: 2026-11-02T00:00:00.000Z  
Trust: Official Source (official_source)

## Why Included

LongCat 2.5 is available through Vercel AI Gateway for coding agents, with text and image input, tool use, and a 1M-token context window.

## Source Summary

**LongCat 2.5** from Meituan is available on Vercel AI Gateway as `meituan/longcat-2.5`. It accepts text and images, supports coding and multi-step tool use, and has a **1M-token context window**.

## Practical Implication

Builders can select it in Claude Code, Codex, Cursor, or another supported agent after installing the latest Vercel CLI and running setup. The listed use cases include visual programming, 3D tasks, and longer workflows from planning through execution.

## Agent-Ready Context

**LongCat 2.5** from Meituan is available on Vercel AI Gateway as `meituan/longcat-2.5`. It accepts text and images, supports coding and multi-step tool use, and has a **1M-token context window**.

Builders can select it in Claude Code, Codex, Cursor, or another supported agent after installing the latest Vercel CLI and running setup. The listed use cases include visual programming, 3D tasks, and longer workflows from planning through execution.

The capability claims come from Meituan; the material provides no benchmark results or comparative measurements. Gateway pricing reflects provider pricing, with no platform fee on inference, including BYOK requests.

## Connected Context

Feed7 judgment across 990 accumulated Signals:

LongCat 2.5 expands the selectable gateway pool with text and image input, coding and multi-step tool-use support, and a 1M-token context window. Meituan's capability claims are not backed here by comparative results, so the listing broadens evaluation options without establishing a model choice.

- [Gemini 3.8 Flash now available on AI Gateway](https://feed7.dev/p/gemini-3-8-flash-now-available-on-ai-gateway-0cun77q) — Both add multimodal, million-token coding-agent routes through AI Gateway without comparative results; the target adds another candidate to evaluate against Gemini on matched workloads.
- [GLM 5.3 now available on AI Gateway](https://feed7.dev/p/glm-5-3-now-available-on-ai-gateway-0s7o9zv) — Both offer a 1M-token gateway route for coding agents, but only the GLM candidate includes a provider claim about long-horizon engineering and token use. Neither supplied record establishes comparative performance.
- [GPT-6.1 Terra now available on AI Gateway](https://feed7.dev/p/gpt-6-1-terra-now-available-on-ai-gateway-1b388mk) — Both support image input and coding agents through AI Gateway, and both lack comparative measurements. LongCat specifies a 1M-token context window, while Terra's listed context is above 1M; these capacity claims alone do not settle model choice.
- [Qwen 3.8 Flash now available on AI Gateway](https://feed7.dev/p/qwen-3-8-flash-now-available-on-ai-gateway-1skoa7y) — Both add text-and-image, million-token routes to the coding-agent pool without benchmark evidence. Qwen specifies a 65K output ceiling, which LongCat's brief does not provide, so that dimension remains uncompared.

## Context Map

- Layer: model
- Domains: coding, image
- Topics: model-selection, coding-agents, reasoning

## Uncertainty

- The capability claims come from Meituan; the material provides no benchmark results or comparative measurements. Gateway pricing reflects provider pricing, with no platform fee on inference, including BYOK requests.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.