GLM-4.7-Flash (Cloudflare)

GLM-4.7-Flash is presented on Cloudflare Workers AI as a lightweight member of the GLM Flash family, surfaced under the catalog identifier that points to the Z.ai-originated lineage. The single available third-party listing characterizes it as a "Free coding model" served through Cloudflare's tier-S deployment, and frames the model as open-weight, aligning with the broader positioning of the GLM Flash line. Because no official Z.ai model card, technical report, or Cloudflare Workers AI model page is included in the supplied evidence, these details should be read as directory-reported framing rather than creator-confirmed specifications, and deeper architecture or training-recipe claims are intentionally omitted here.

The model is positioned for code-generation and software-engineering tasks, with a SWE-bench score of roughly 59.2% reported by the third-party directory that places it in a strong performance tier for its class. Its 131,000-token context window enables long-file reasoning and multi-turn coding assistance, which makes it suitable for repository-scale understanding, refactoring assistance, and agent-driven development workflows. As an open-weight model hosted on Cloudflare Workers AI without a credit-card requirement, GLM-4.7-Flash offers developers a low-friction path to experiment with a code-focused large language model, though commercial-use terms remain unverified and prospective users should confirm licensing before production deployment.