GLM-5.2
GLM-5.2 is an open-weights language model from Z.ai, released June 16, 2026. It has 753B parameters (mixture-of-experts, 51B active). The context window is 1M tokens. On the Epoch Capabilities Index it scores 153.0, rank 54 of 256 rated models. API price: $1.4 input / $4.4 output per million tokens.
We're introducing GLM-5.2, our latest flagship model for long-horizon tasks. It marks a substantial leap in long-horizon task capability over its predecessor GLM-5.1 and, for the first time, delivers that capability on a solid 1M-token context.
Check it against your own PC →
Can you run GLM-5.2 locally?
Even with 32 GB of VRAM and 32 GB of RAM it is too large to run at a usable speed on a single consumer GPU. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Too large | – | – | – | – |
| 8 GB VRAM | Too large | – | – | – | – |
| 12 GB VRAM | Too large | – | – | – | – |
| 16 GB VRAM | Too large | – | – | – | – |
| 24 GB VRAM | Too large | – | – | – | – |
| 32 GB VRAM | Too large | – | – | – | – |
Ready-made quantizations on Hugging Face: unsloth/GLM-5.2-GGUF · huihui-ai/Huihui-GLM-5.2-abliterated-GGUF · DevQuasar/zai-org.GLM-5.2-GGUF · turintech/GLM-5.2-MaxMin108-GGUF · brandonmusic/GLM-5.2-EXL3-TR3v4-3.5bpw-MTP78 · malaiwah/GLM-5.2-EXL3-FQ-segments
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 15.2 | Chess Puzzles, SimpleBench, Mystery Game Puzzles |
| Math | 46.1 | ProofBench, OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private |
| Coding | 51.1 | MirrorCode, SWE-bench Verified, DeepSWE |
| Knowledge | 71.9 | GPQA diamond, SimpleQA Verified |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| GPQA Diamond | 92.9% |
| OTIS Mock AIME | 86.4% |
| SWE-bench Verified | 78.7% |
| FrontierMath (Tiers 1–3) | 62.3% |
| SimpleBench | 50.6% |
| DeepSWE | 43.8% |
| MirrorCode | 36.7% |
| ProofBench | 35.0% |
| SimpleQA Verified | 34.2% |
| FrontierMath-Tier-4-v2-Private | 29.3% |
| Chess Puzzles | 17.9% |
| Mystery Game Puzzles | 10.8% |
| EBR-bench | 9.5% |
Links
Announcement · Hugging Face: zai-org/GLM-5.2 · Epoch AI benchmarks
Models with a similar score
Claude Sonnet 5 · DeepSeek V4.1 Flash · GPT-5.6 Luna · GPT-6 Luna · Claude Opus 4.6 · Grok 4.7 · Gemini 3.1 Pro · Qwen3.8 Max (0902)
More from Z.ai
GLM-5.3-FlashX · GLM 5.3 Flash BF16 · GLM 5.3 BF16 · GLM-5.3-Flash · GLM-5.3 · GLM-5.1 · GLM-5V-Turbo · GLM-5-Turbo