Kolors
Kwai's Apache-2.0 bilingual text-to-image model — solid Chinese/English generation, open for commercial use.
About
Kolors (July 2024) is Kwai (Kuaishou)'s bilingual Chinese/English text-to-image model, built on a DiT architecture with a GLM text encoder. At release it was among the strongest open models for Chinese-language prompts.
Apache 2.0 licensed — commercial use allowed. The caveat is age: a mid-2024 model, it has been superseded on quality by 2025-2026 releases (Z-Image, FLUX.2, Qwen-Image) and shows minimal recent development. Direct Hugging Face adoption is small (828 downloads on the canonical checkpoint); historical usage concentrated on GitHub and ModelScope. Reasonable as a known-quantity Apache-2.0 baseline, but new projects should evaluate newer alternatives first.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-claimed at release (2024): competitive bilingual quality — treat as provider-claimed and dated. No current leaderboard presence.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 5/10
Scored 5/10 (metrics tier). Apache 2.0 with genuine bilingual strength at release, but this is a 2024 model in a 2026 world: (1) superseded on quality by Z-Image, FLUX.2, and Qwen-Image; (2) minimal recent development or community momentum; (3) small direct adoption footprint; (4) quality claims are dated provider marketing. A historical Apache-2.0 baseline, not a current pick.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 4.8/10
- Maintenance · recency, releases, issue load
- 4.0/10
- Openness · license + docs
- 8.8/10
- Manual review · sanity read + red-flag check
- 6.0/10
- · HF downloads 828 (peer-normalized)
- · released 823 days ago
- · manual review: 2024 model superseded by Z-Image/FLUX.2/Qwen-Image; minimal recent development; license: Apache-2.0