

GLM 5.1
#31 in Frontier LLMszhipu-ai · v5.1 · since 2026-04-07 · 85× · last seen Jun 30, 2026
GLM-5.1 (also known as GLM-Z1-Rumination) is the fifth-generation frontier language model from Z.ai (formerly Zhipu AI), released on April 7/8, 2026. It is built on a Mixture-of-Experts architecture with 744 billion total parameters (40 billion active per token) and represents a post-training upgrade over the GLM-5 base model. The model is specifically designed for long-horizon autonomous engineering tasks and can, according to the vendor, work continuously on a single task for up to 8 hours without human intervention. It was released as an open-weight model under the MIT license on Hugging Face and topped the SWE-Bench Pro leaderboard at launch.
Features
| Key Benchmark (%) | SWE-Bench Pro: 58.4% (SOTA at launch, ahead of GPT-5.4 at 57.7% and Claude Opus 4.6 at 57.3%); AIME 2026: 95.3%; GPQA-Diamond: 86.2%; Artificial Analysis Intelligence Index: 40 points |
| Context Window (Tokens) | 200,000 tokens (context input); max. 128,000 tokens output |
| License | MIT License (open-weight; commercial use, modification, and redistribution permitted) |
| Multimodality | Text only (text-in, text-out); no image or audio input. For multimodal tasks, Z.ai recommends the separate GLM-5V-Turbo model. |
| Price | GLM Coding Plan: Lite ~$54/quarter, Pro ~$216/quarter, Max ~$480/quarter (all tiers incl. GLM-5.1 access); pay-as-you-go API via third-party providers from $0.975/1M input tokens |
| Release Date | April 7, 2026 (API: March 27, 2026; open-source weights: April 7/8, 2026) |