ChatGLM3 Models by Tsinghua Knowledge Engineering Group (THUDM)
About
The ChatGLM3 family, created by Zhipu AI and the Knowledge Engineering Group at Tsinghua University, includes a series of open-source large language models specially crafted for dialogue applications. These models advance the capabilities of prior versions, notably improving dialogue fluency and ease of deployment. One standout model within this family is ChatGLM3-6B-Base, which excels in performance among models with fewer than 10 billion parameters, evaluated across various benchmarks. Key features of the ChatGLM3 models include advanced functionality like tool invocation, code execution, and support for complex agent tasks. The family comprises several variants, such as ChatGLM3-6B and longer-context versions like ChatGLM3-6B-32K and ChatGLM3-6B-128K, all accessible for academic research and available for commercial use without charge upon registration. The models are fully available on platforms like Hugging Face and GitHub 12.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 8k context and 6B parameters.
Use when the workload needs 32k context and 6B parameters.
Use when the workload needs 128k context and 6B parameters.
Use when the workload needs 131k context and 6.2 parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| ChatGLM3 6B | Use when the workload needs 8k context and 6B parameters. | 2024-01 | 8k context6B parameters | Current |
| ChatGLM3 6B 32K | Use when the workload needs 32k context and 6B parameters. | 2024-01 | 32k context6B parameters | Current |
| ChatGLM3 6B 128K | Use when the workload needs 128k context and 6B parameters. | 2024-01 | 128k context6B parameters | Current |
| ChatGLM 3 6B | Use when the workload needs 131k context and 6.2 parameters. | 2023-06 | 131k context6.2 parameters | Current |
Release Timeline
2 release groupsSpecifications(4 models)
| Model | Released | Context | Parameters |
|---|---|---|---|
| ChatGLM3 6B | 2024-01 | 8k | 6B |
| ChatGLM3 6B 32K | 2024-01 | 32k | 6B |
| ChatGLM3 6B 128K | 2024-01 | 128k | 6B |
| ChatGLM 3 6B | 2023-06 | 131k | 6.2 |
Available From(2 providers)
Frequently Asked Questions
- What is ChatGLM3 used for?
- ChatGLM3 is used for coding, agent workflows, and long-context generation. The family description and listed model capabilities point to those workloads as the best fit.
- How does ChatGLM3 compare to ChatGLM-4?
- ChatGLM3 by Tsinghua Knowledge Engineering Group (THUDM) is strongest where you need coding, while ChatGLM-4 by Tsinghua Knowledge Engineering Group (THUDM) is the closest related family to check for adjacent model selection. ChatGLM3 has 4 listed variants and reaches up to 131k context, while ChatGLM-4 reaches up to 128k context, so compare the specs and pricing tables before choosing a production model.
- Which ChatGLM3 model should I use?
- If price is the main constraint, use the pricing table first because ChatGLM3 does not have complete provider pricing in the local data. For the most capable/latest local choice, evaluate ChatGLM 3 6B with 131k context.
