About

The consultancy born at the intersection of behavioral economics and human experience.

René Studio
AI EXPERIENCE STUDIO

The CX design platform we built from a decade of client work.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

Mystery audits we run, scored live for every location.

Explore Mysterya
NOW HIRING

Join a team reshaping how the world experiences brands.

View open roles →

COMPANY

GROW WITH US

CONNECT

Services

Comprehensive CX and management consulting for enterprise brands.

René Studio
AI EXPERIENCE STUDIO

Every engagement, mapped and scored in one AI workspace.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

Mystery shopping, scored and evidenced in real time.

Explore Mysterya
ALL SERVICES

Explore the full range of CX & management consulting services.

Browse all services →

CORE

SPECIALIST

Solutions

Structured solutions that turn CX ambition into measurable outcomes.

René Studio
AI EXPERIENCE STUDIO

Map, score and fix the journeys we redesign, with AI.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

Check the experience you designed is delivered.

Explore Mysterya
ALL SOLUTIONS

Explore every CX solution we offer.

Browse solutions →

STRATEGY & GOVERNANCE

DESIGN & DELIVERY

CULTURE & EXPERIENCE

Industries

A decade of CX transformation across the region's defining sectors.

René Studio
AI EXPERIENCE STUDIO

Sector-ready journeys, scored by AI in minutes.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

Mystery audits tuned to your sector, scored live.

Explore Mysterya
ALL INDUSTRIES

See how we work across every sector.

Browse industries →

BUILT ENVIRONMENT

FINANCE & TECH

PEOPLE & MOBILITY

Products

Proprietary tools, platforms, and AI that power CX transformation.

René Studio
AI EXPERIENCE STUDIO

Design, score and fix customer journeys with AI.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

Audit every location and act on what you find.

Explore Mysterya
REBELDECK A · 36 FORCES

The forces that shape how humans experience the world.

Explore REBEL Reveal →
ALL PRODUCTS

Explore the full Renascence product ecosystem.

Browse products →

PLATFORMS

AI & TECHNOLOGY

LEARNING & FRAMEWORKS

CX TOOLKIT

Opinion

Insights, research, and conversations at the frontier of CX.

René Studio
AI EXPERIENCE STUDIO

Turn what you read into a journey you can score.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

See what customers really experience, visit by visit.

Explore Mysterya
ReadExperience JournalArticles & research on CX, behavior, and transformation.Watch & listenExperience LoomOur video podcast on CX & behavior.CuratedCX NewsIndustry news that matters in CX, minus the noise.

Latest articles

Latest episodes

Latest news

Hub

Free tools, templates, and resources to advance your CX practice.

René Studio
AI EXPERIENCE STUDIO

Design, score and fix customer journeys with AI.

Explore René Studio
Mysterya
MYSTERY AUDITS, LIVE

Every visit, every location, every score. Live.

Explore Mysterya
THE MANIFESTOBurn the Deck.
Ten Virtues. Zero Excuses.Start reading →
THE HUB

Every free tool, template and resource in one place.

Visit the Hub →

AI TOOLS

FREE TOOLS

LEARNING

CULTURE

AI · 17 November 2026

Meituan's LongCat-2.5 agent model: 1.6T parameters on Huawei Ascend

Meituan has released model weights for LongCat-2.5, a 1.6-trillion-parameter agent model that scores 88.4 on Terminal Bench, just behind Claude Opus 5, according to reporting by The Decoder. The sample deployment runs on 194 Huawei Ascend chips.

Newsdesk
Curated briefing · 2 min read

What happened

Chinese technology group Meituan has released the model weights for LongCat-2.5, a 1.6-trillion-parameter model built for AI agents and, according to its reporting, trained entirely on Chinese hardware. The release was covered by The Decoder, which framed it around the model's hardware sourcing.

On Terminal Bench, a benchmark used to assess agent performance, LongCat-2.5 scores 88.4 points. That places it just behind Claude Opus 5, the model developed by Anthropic, according to the same report.

The sample deployment described in the coverage runs on 194 Huawei Ascend chips. Meituan's material, as reported, makes no reference to Nvidia GPUs, which is the focus of the headline the story has carried across outlets.

Why it matters

The significance here lies less in a single benchmark score than in what the release signals about agent-focused AI. A model of this scale aimed squarely at agentic tasks, meaning systems that plan and carry out multi-step work rather than simply answering prompts, changes what organisations can build on. Leaders evaluating AI platforms will need to weigh capability alongside where and how a model runs.

For digital transformation teams, the deployment detail matters as much as the benchmark. A model demonstrated on a specific accelerator stack raises practical questions about infrastructure choices, procurement, portability and the cost of running agents at scale. Operators in any market will want to understand which parts of their AI roadmap depend on particular hardware, and how quickly that dependency can change.

By the numbers

  • 1.6 trillion parameters in the LongCat-2.5 model.
  • 88.4 points on Terminal Bench, just behind Claude Opus 5.
  • 194 Huawei Ascend chips used in the sample deployment.

The Renascence take

Most of the coverage is fixated on the hardware story, but the experience question is more useful. An agent is only as good as the service it delivers in the moments that matter, and benchmark rankings say little about whether a customer's problem actually gets resolved. The behavioural principle underneath is simple: people judge a system by its failures and recoveries, not its leaderboard position.

Benchmarks reward capability, but customers reward reliability, clarity and a graceful handover when the agent gets stuck. The real test of an agent model is not whether it lands near the top of a table, but whether it keeps a customer's trust through a messy, multi-step request. A customer-obsessed operator should therefore run pilots on their own journeys, measure task completion and escalation quality, and treat hardware choices as a service-continuity decision rather than a headline.

Sources

This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.

FAQ

Questions we get on this topic

LongCat-2.5 is a 1.6-trillion-parameter model from Chinese technology group Meituan, built for AI agents, meaning systems that plan and carry out multi-step work. Meituan released its model weights, according to reporting by The Decoder.

LongCat-2.5 scores 88.4 points on Terminal Bench, a benchmark used to assess agent performance. According to the same reporting, that places it just behind Claude Opus 5, developed by Anthropic.

Coverage describes a sample deployment running on 194 Huawei Ascend chips. Meituan's material, as reported, makes no reference to Nvidia GPUs, and the source reporting says the model was trained entirely on Chinese hardware.

A model demonstrated on a specific accelerator stack raises questions about infrastructure, procurement, portability and the cost of running agents at scale. Renascence suggests operators pilot agents on their own customer journeys and measure task completion and escalation quality.

Stay ahead of CX

Get the signal, not the noise.

The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.