Openness profile
Nex AGI Nex (Nex-N2.6; the N1 and N2.x line)
Apache-2.0 post-training of another lab's open base, where the openness stops at the checkpoint.
The six questions
Read the rubric in Chapter 13Deployment
2 of 3 · Yes, with caveats
SGLang/vLLM serving and third-party GGUF quants; a 328 GB FP8-dominant checkpoint aimed at datacenter hardware
Deployment in the bookTransparency
1 of 3 · Limited / partial
the base model is named, but the post-training data, code and recipe are undisclosed, and the published evals are vendor-reported
Transparency in the bookGovernance & Participation
1 of 3 · Cosmetic participation
An institute-initiated alliance with hidden GitHub membership, no governance charter, no public issue tracker and no published decision process; the legal operator is a Shanghai company with no disclosed funding or leadership.
Governance in the bookWhat is actually released
- weights (Nex-N2.6, ~321B total / ~18B active sparse MoE post-trained on GLM-5.3-Flash; 129 FP8 shards, ~328 GB)
- Glm5NextForConditionalGeneration modeling code
- GitHub mirror and SGLang usage instructions
- third-party GGUF and quantized builds
- vendor-reported benchmark tables (OSWorld, Terminal-Bench 2.1, AutomationBench); no training code, no corpus, no paper for N2.6
Terms of use
Artifact license
Apache License 2.0 (HF license:apache-2.0 tag on Nex-N2.6; consistent across the N1 and N2.x releases)
Read the license source huggingface.coArtifact license: OSI-approved
OSI approval here describes the listed artifact license. It does not certify the full release under the Open Source AI Definition.
Separate access agreements
No additional download or access agreements are recorded for this assessment. Check the cited release terms.
Download gates and accompanying contracts are recorded separately from the artifact license’s approval status.
Assessment record
- Assessment scope
- Only the named artifacts and release versions listed in this profile; no score applies to other releases from this publisher.
- Artifact kind
- Trained model
- Assessment recorded
Tier rationale
Permission/deployment gate: P<2 or D<3. A=3 P=3 D=2 T=1 R=1. G is displayed separately.
How the tier is assigned →What the publisher claims
agentic post-training built on other labs' open bases
open-source, open ecosystem
Why it matters
Nex AGI markets an open ecosystem while publishing only weights: no data, no code, no recipe, no governance. It shows how an open license on a borrowed base can carry an openness reputation the artifacts do not.
Sources
- nex-agi/Nex-N2.6 huggingface.co
- sii-nex/Nex-N2.6 github.com
- Website nex-agi.com
- nex-agi/nex-n2.6 openrouter.ai
- abs/2512.04987 arxiv.org
- hub/companies/nex-agi hokai.io
Dataset snapshot 2026-11-15 · Source and history github.com . An assessment date records this profile; it does not establish that every source was rechecked that day. Scores are qualitative readings of the cited evidence; the tier follows the published rules and does not measure intelligence.