GLM-5.1 Launches: 94.6% of Claude Opus 4.6 Coding Performance at $3/Month

Released: March 27, 2026 | Source: Apiyi.com

Z.ai (formerly Zhipu AI) quietly dropped GLM-5.1 yesterday — and the coding benchmark numbers are turning heads.

The numbers

• Scored 45.3 on Claude Code's coding evaluation
• Claude Opus 4.6 scores 47.9 — GLM-5.1 hits 94.6% of that
• 28% improvement over GLM-5 (35.4 → 45.3) in just ~6 weeks
• Available now to all GLM Coding Plan users (Lite/Pro/Max)

Pricing

• Promotional: $3/month
• Standard: $10/month
• Switch by changing model name to glm-5.1 in your API call

The kicker: this evaluation used Claude Code as the testing framework — meaning Claude's own tooling naturally favors Claude. GLM-5.1 nearly matched Opus anyway.

Bottom line: If Claude Opus costs are eating your budget, GLM-5.1 is now the most credible alternative for coding use cases.

→ Full breakdown and API guide

Subscribe to LGTM for daily AI model news delivered to your inbox.