Released 2026-07-02 · general agent
Deployment evidence
Meituan · 1.6T total · approximately 48B active · 1M tokens
Best-fit scenarios
Long-horizon coding, search, repository edits and tool-driven agents on GPU or NPU clusters.
Not recommended
- Workstation deployment
- Quoting in-house benchmark results as ChinaAPI reproduction
Repository-scale coding
Vendor-statedOfficial materials emphasize repository edits and integrations with coding-agent harnesses.
Capability sourceSearch and general agents
Public evidenceThe vendor publishes coding, BrowseComp, RWSearch and agent evaluations.
Capability sourceNPU-based sovereign deployment
Vendor-statedAn official SGLang-FluentLLM NPU serving path is linked.
Capability sourceOfficial launch / minimum
Evidence CNot published
GPU and NPU paths are documented, but the vendor does not publish an absolute minimum hardware shape.
200-person team
Evidence ECluster sizing required
A 1.6T model needs a measured cluster design; employee count alone is not a capacity input.
Commercial API
Evidence COfficial serving paths
Official GPU and NPU serving paths exist; topology, redundancy and throughput remain operator-specific. Primary recipe
Commercial-use check
MIT
MaaS / hosted service: No model-specific MaaS restriction identified in the MIT license.
Attribution: Retain the copyright and permission notice.
Known limitations and open questions
- No official minimum GPU count
- Most published evaluation values are vendor-measured