Rankings / Alibaba (Qwen)
Qwen3-235B-A22B-Instruct (open weights, self-hosted reference)
Usage share 0.13% · OpenRouter rankings API (daily token share, 2026-08-04)
Missing for Tier 2: external pre-deployment testing, third-party certification. The tier is computed from this checklist — satisfying these requirements moves the badge, automatically.
What happens to your data: training-on-customer-data defaults, retention windows, residency options, and the enterprise-versus-consumer terms gap. A legal property, not a capability — it survives every model generation.
Deployer property: in the self-hosted reference deployment, data never leaves deployer infrastructure. Alibaba Cloud's hosted Model Studio has separate terms not graded here.
Whether it changes without warning: versioning discipline, changelog quality, deprecation policy, and observed silent changes. The signal no one else tracks.
Immutable published checkpoints with explicit versioned releases; no silent-change surface in the self-hosted reference deployment.
Whether an attacker can make it misbehave — direct jailbreaks against the model's own policies and indirect prompt injection in agentic tool use. Graded to the weaker of the two, because an attacker takes the easier path.
Independent testing shows high attack success on the bare weights; safety behavior is light and easily bypassed, consistent with most open-weight releases.
No injection hardening or published agent-scenario safety results.
Whether you can see how it was built and tested: model cards, published safety evals, external pre-deployment testing, and disclosure of changes. The mechanism behind the tier ladder.
Detailed technical report and model card, Apache 2.0 license; no safety evals of substance and no external testing.
“While Qwen2.5 was pre-trained on 18 trillion tokens, Qwen3 uses nearly twice that amount, with approximately 36 trillion tokens covering 119 languages and dialects.”
Whether it is certified and compliant: SOC 2, ISO/IEC 42001, HIPAA eligibility, EU AI Act readiness, and audit availability.
Deployer property: no certification attaches to the weights; deployer inherits all compliance obligations.
No regional pinning — the provider chooses where data is processed.
Self-hosted reference deployment; Alibaba Cloud's hosted Model Studio has separate terms not graded here.
No consumer tier exists for the self-hosted reference deployment; data handling is deployer-controlled.
Level tiers can mean both are clean, or that the API tier is itself weak with nothing better to compare against.
One tracked change, on 2026-08-04 — 1 days before the as-of date (2026-08-05). A single event cannot establish a cadence, so days-since is shown without a baseline.
No dated score readings recorded for this model yet. Readings are only entered where multiple real, dated third-party values exist — never interpolated.
retrieved 2026-08-03 — 2026-08-05
Tiers are now derived from a per-requirement checklist rather than assigned, and not-applicable / under-review became first-class grade states excluded from the composite score. Deployer-property vectors on open-weight models (data handling, compliance) moved from graded to not-applicable, changing the composite scores of the tagged models.
Compare this model: Qwen3 235B-A22B + open compare view →