LongCat-2.5: Meituan's 1.6T multimodal model for long-horizon coding agents
Meituan released LongCat-2.5 on 2 November, a native multimodal model with 1.6 trillion total parameters and about 48 billion activated per token, trained for long-horizon coding and agent tasks and available via chat and API, with GitHub and Hugging Face releases planned.
Why it matters
Trained on extended agent trajectories, including failed attempts turned into reproducible training tasks, and reports 88.4 on Terminal Bench v2.1; built for multi-action work across whole codebases, not single-turn code completion.
What you could build with it
An indie developer-tools startup could build a long-running codebase migration service on LongCat-2.5: point it at a legacy repository, let it work through refactors across many files over hours, and charge per completed migration with human review at checkpoints.
Does it hold up?
Too early to judge: only chat and API access are live so far, with GitHub and Hugging Face releases still planned, so no independent runs exist yet.
Built with LongCat-2.5
- Meituan launches LongCat-2.5 for extended coding tasksTechNode · Release report: 1.6T total parameters, ~48B active per token, unified visual-text tokens, chat and API live with GitHub/Hugging Face releases planned.
- Meituan releases LongCat-2.5, a 1.6 trillion parameter model for agentic codingDaily AI Brief · Coverage of the 2 November release, its long-horizon coding focus, Claude Code/Codex/Hermes tool support, and benchmark claims of 88.4 on Terminal Bench v2.1.
- Meituan releases LongCat-2.5 for long-running coding tasksAI Affairs · Notes the September API preview lineage, low-precision inference optimised for domestic chips, and that the 1.6T parameter figure is disputed in some reports.
Learn more
First spotted on article: source.
More AI models
VoiceStudio: open-source, fully-local ElevenLabs alternative for voice cloning, dubbing, and audiobooksdebpalash/VoiceStudio is a desktop app that runs voice cloning, voice design, video dubbing, dictation, transcription…models · JEV 0.75Tencent Hunyuan Hy3: open 295B MoE with 21B active under Apache 2.0Tencent open-sourced Hy3, a 295B-parameter mixture-of-experts model with only 21B parameters active per token, under…models · JEV 0.73NVIDIA Isaac Lab 3.0 GA + Isaac OS open runtime + Agile ONE SN humanoidMajor Isaac platform expansion at GTC Berlin Oct 21: Isaac Lab 3.0 is generally available as the open-source robot…models · JEV 0.73NVIDIA Nemotron 3.5 Retriever collection: Embed 8B/1B, Rerank 8B, Parse 631MNVIDIA's Oct 21 collection drop for agentic retrieval: Nemotron-3.5-Embed in 8B BF16, 1B BF16 and 1B NVFP4…models · JEV 0.72Kimi K3.1 Pro: Moonshot's 2.8T open-weight flagship officially launchesMoonshot AI officially launched Kimi K3.1 Pro on Oct 23, a mixture-of-experts model with 2.8 trillion total / 104…models · JEV 0.72NVIDIA Cosmos 3.5 world models: now openly availableNVIDIA's Cosmos 3.5 family of open world foundation models for robots, autonomous vehicles and vision agents is now…models · JEV 0.71
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.