Zhipu AI model product
GLM 5.2
AI's flagship foundation model built for long-horizon tasks, delivering a solid 1M-token context that stably sustains long, messy coding-agent trajectories.
Updated Aug 12, 2026. Default version: GLM-5.2
LLMBoard score79.3GLM-5.2
Coverage80%19 benchmark families
Context window1MTokens
Official input price$1.4Zhipu AI API
GLM 5.2 Specifications
Technical details for the model's default version.
- Version
- GLM-5.2
- Released
- Jun 16, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 753B
- Context window
- 1M
- Max output
- 131.1K
- Inputs
- text
- Outputs
- text
- Open weights
- Yes
- License
- MIT
GLM 5.2 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
GLM 5.2 Benchmark Results
Benchmark scores for GLM-5.2.
| AIME 2026 | 99.2% | 01 | 18 | 100.0% | C | |
| CritPT | 16.7% | 01 | 4 | 100.0% | C | |
| IMO-AnswerBench | 91.0% | 02 | 19 | 94.4% | C | |
| Program Bench | 63.7% | 02 | 5 | 75.0% | C | |
| NL2Repo | 48.9% | 03 | 14 | 84.6% | C | |
| PostTrainBench | 34.3% | 03 | 5 | 50.0% | C | |
| SWE-Marathon | 13.0% | 03 | 3 | 0.0% | C | |
| FrontierSWE | 74.0% | 04 | 15 | 78.6% | B | |
| HMMT Feb 26 | 92.5% | 06 | 11 | 50.0% | C | |
| DeepSWE | 46.2% | 07 | 10 | 33.3% | C | |
| HMMT 2025 | 94.4% | 09 | 33 | 75.0% | C | |
| MCP Atlas | 76.8% | 09 | 31 | 73.3% | C | |
| Terminal-Bench 2.1 | 82.7% | 10 | 19 | 50.0% | C | |
| SWE-Bench Pro | 62.1% | 11 | 45 | 77.3% | C | |
| FrontierCode 1.1 | 24.5% | 12 | 15 | 21.4% | B | |
| Humanity's Last Exam | 54.7% | 12 | 93 | 88.0% | C | |
| DeepSWE 1.1 | 44.0% | 16 | 20 | 21.1% | B | |
| GPQA | 91.2% | 18 | 234 | 92.7% | C | |
| Toolathlon | 48.2% | 18 | 31 | 43.3% | C |
GLM 5.2 Arena Results
Preference and agent-evaluation results for the default version.
| agent task outcome explicit | overall | 05 | 0.1 | N/A | 28.6K | |
| webdev | overall | 07 | 1588.2 | 7,764 | N/A | |
| agent praise complaint | overall | 10 | 0.1 | N/A | 9.6K | |
| agent | overall | 12 | 0.1 | N/A | 2.1M | |
| agent tool hallucination | overall | 12 | 0.0 | N/A | 2M | |
| agent steerability | overall | 15 | 0.1 | N/A | 38.8K | |
| agent bash recovery steps | overall | 23 | 0.1 | N/A | 60.2K | |
| text | overall | 26 | 1464.6 | 26,370 | N/A | |
| text factuality | overall | 33 | 1459.0 | 26,349 | N/A | |
| text style control | overall | 33 | 1469.8 | 26,370 | N/A |
GLM 5.2 Runtime Performance
Provider-specific output speed and catalog latency for GLM-5.2. Runtime does not affect the capability score.
| DeepInfra | 87.154 tok/s | 1.199 s | 1M | 131.1K | |
| Together | 47.813 tok/s | 5.91 s | 1M | 131.1K | |
| Fireworks | 45.606 tok/s | 1.068 s | 1M | 131.1K | |
| Novita | 41.389 tok/s | 7.464 s | 1M | 131.1K | |
| FriendliAI | 16.583 tok/s | 0.596 s | 1M | 131.1K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
GLM 5.2 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $1.4 input, $4.4 output per 1M
- Official provider
- Zhipu AI
- Lowest third-party
- From $0.42 input, $1.32 output per 1M via NanoGPT
- Tracked offerings
- 77
| Nvidia | z-ai/glm-5.2 | global | N/A | N/A | 1M | |
| Zhipu AI Coding Plan | glm-5.2 | global | N/A | N/A | 1M | |
| Kenari | glm-5-2 | global | N/A | N/A | 1M | |
| UnoRouter | glm-5.2:free | global | N/A | N/A | 1M | |
| Alibaba Token Plan | glm-5.2 | global | N/A | N/A | 1M | |
| Umans AI Coding Plan | umans-glm-5.2 | global | N/A | N/A | 405.5K | |
| Z.AI Coding Plan | glm-5.2 | global | N/A | N/A | 1M | |
| Alibaba Token Plan (China) | glm-5.2 | global | N/A | N/A | 1M | |
| NanoGPT | zai-org/glm-5.2 | global | $0.42 | $1.32 | 1M | |
| CrofAI | glm-5.2 | global | $0.50 | $2.2 | 1M | |
| LLM Gateway | glm-5.2 | global | $0.55 | $1.93 | 1M | |
| Deep Infra | zai-org/GLM-5.2 | global | $0.75 | $2.4 | 1M | |
| Weights & Biases | zai-org/GLM-5.2 | global | $0.76 | $2.42 | 262.1K | |
| OpenRouter | z-ai/glm-5.2 | global | $0.76 | $2.42 | 1M | |
| routing.run | glm-5.2 | global | $0.80 | $2.4 | 200K | |
| Vultr | zai-org/GLM-5.2-FP8 | global | $0.85 | $3.1 | 393.2K | |
| Lilac | zai-org/glm-5.2 | global | $0.90 | $3 | 524.3K | |
| GMI Cloud | zai-org/GLM-5.2-FP8 | global | $0.979 | $3.08 | 1M | |
| ai& | zai-org/glm-5.2 | global | $1 | $4 | 1M | |
| Ambient | z-ai/glm-5.2 | global | $1.05 | $4.4 | 202.8K | |
| Merge Gateway | zai/glm-5.2 | global | $1.05 | $3.3 | 1M | |
| Vercel AI Gateway | zai/glm-5.2 | global | $1.1 | $3.85 | 1M | |
| Alibaba (China) | glm-5.2 | global | $1.1 | $3.85 | 1M | |
| AIHubMix | glm-5.2 | global | $1.13 | $3.94 | 1M | |
| CrossModel | z-ai/glm-5.2 | global | $1.2 | $4.4 | 1M | |
| Vivgrid | glm-5.2 | global | $1.2 | $4.2 | 1M | |
| Ambient | zai-org/GLM-5.2-FP8 | global | $1.2 | $4.2 | 202.8K | |
| Inceptron | zai-org/GLM-5.2 | global | $1.2 | $4.2 | 1M | |
| Wafer | GLM-5.2 | global | $1.2 | $4.1 | 1M | |
| Requesty | glm-5.2 | global | $1.2 | $4.2 | 1M | |
| Cortecs | glm-5.2 | global | $1.2 | $4.2 | 1M | |
| DInference | glm-5.2 | global | $1.25 | $3.89 | 1M | |
| GreenPT | glm-5.2 | global | $1.25 | $5.02 | 1M | |
| Zhipu AI | glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Impossibl | fireworks/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Impossibl | zai/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Fireworks AI | accounts/fireworks/models/glm-5p2 | global | $1.4 | $4.4 | 1M | |
| OpenCode Go | glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Baseten | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 1M | |
| Nebius Token Factory | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 432K | |
| Friendli | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 1M | |
| OpenCode Zen | glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Cloudflare Workers AI | @cf/zai-org/glm-5.2 | global | $1.4 | $4.4 | 262.1K | |
| Neon | glm-5-2 | global | $1.4 | $4.4 | 1M | |
| ClinePass | cline-pass/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Pioneer | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 1M | |
| SiliconFlow | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 1M | |
| ZenMux | z-ai/glm-5.2 | global | $1.4 | $4.5 | 1M | |
| Abacus | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 1M | |
| EmpirioLabs AI | glm-5-2 | global | $1.4 | $4.4 | 1M | |
| Venice AI | zai-org-glm-5-2 | global | $1.4 | $4.4 | 1M | |
| Together AI | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 512K | |
| DigitalOcean | glm-5.2 | global | $1.4 | $4.4 | 262.1K | |
| Umans AI | umans-glm-5.2 | global | $1.4 | $4.4 | 405.5K | |
| Ofox | z-ai/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Synthetic | hf:zai-org/GLM-5.2 | global | $1.4 | $4.4 | 524.3K | |
| Kilo Gateway | z-ai/glm-5.2 | global | $1.4 | $4.4 | 262.1K | |
| Subconscious | subconscious/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Charm Hyper | glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Alibaba | glm-5.2 | global | $1.4 | $4.4 | 1M | |
| NovitaAI | zai-org/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Hugging Face | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 262.1K | |
| SiliconFlow (China) | zai-org/GLM-5.2 | global | $1.4 | $4.4 | 1M | |
| Cloudflare AI Gateway | workers-ai/@cf/zai-org/glm-5.2 | global | $1.4 | $4.4 | 262.1K | |
| Z.AI | glm-5.2 | global | $1.4 | $4.4 | 1M | |
| HPC-AI | zai-org/glm-5.2 | global | $1.4 | $4.4 | 1M | |
| Databricks | databricks-glm-5-2 | global | $1.4 | $4.4 | 1M | |
| evroc | zai-org/GLM-5.2 | global | $1.44 | $5.75 | 524.3K | |
| Neuralwatt | glm-5.2 | global | $1.45 | $4.5 | 1M | |
| TensorX | z-ai/glm-5.2 | global | $1.5 | $4.5 | 1M | |
| Tinfoil | glm-5-2 | global | $1.5 | $5.25 | 384K | |
| Berget.AI | zai-org/GLM-5.2 | global | $1.54 | $4.84 | 524.3K | |
| UnoRouter | glm-5.2 | global | $1.6 | $5.03 | 1M | |
| Scaleway | glm-5.2 | global | $1.8 | $5.5 | 256K | |
| Regolo AI | glm5.2 | global | $2.31 | $6 | 96K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
GLM 5.2 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| GLM-5.2 | 79.3 | 753B | 1M | 131.1K | Yes | MIT |
What is GLM 5.2?
Key information about GLM 5.2 and its available data.
AI's flagship foundation model built for long-horizon tasks, delivering a solid 1M-token context that stably sustains long, messy coding-agent trajectories. 1 on SWE-bench Pro) and the highest-ranked open-source model across long-horizon coding benchmarks (FrontierSWE, PostTrainBench, SWE-Marathon).
It introduces flexible thinking effort levels (High and Max) to balance capability against latency and compute. 9x at 1M context, and an improved MTP layer for speculative decoding that raises acceptance length by up to 20%.
Released under a pure MIT open-source license with weights available on HuggingFace and ModelScope, it supports transformers, vLLM, SGLang, xLLM, and ktransformers, with 1M input context, 128K max output, thinking mode, function calling, structured output, context caching, and MCP integration.
Data as of 2026-08-11.
GLM 5.2 vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to GLM 5.2
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
ZA
GLM 5.1
Zhipu AI
ZA
GLM 5
Zhipu AI
ZA
GLM 4.7
Zhipu AI
ZA
GLM 5V Turbo
Zhipu AI
ZA
GLM 4.6
Zhipu AI
ZA
GLM 4.5
Zhipu AI
FAQ
Common questions about GLM 5.2.
When was GLM 5.2 released?
GLM 5.2's default version was released on Jun 16, 2026.
How much does GLM 5.2 cost?
GLM 5.2's official API price is $1.4 per million input tokens and $4.4 per million output tokens via Zhipu AI. The lowest tracked third-party offer starts at $0.42 input and $1.32 output via NanoGPT.
Who created GLM 5.2?
GLM 5.2 was created by Zhipu AI.
What is the context window for GLM 5.2?
The default version has a 1M token context window.
Is GLM 5.2 open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer GLM 5.2?
77 provider offerings are linked to the default version.
What models should I compare GLM 5.2 with?
Nearby ranked alternatives include Grok 4.5, Claude Opus 4.6, Gemini 3.6 Flash.