THUDM: GLM Z1 Rumination 32B

Text input Text output Unavailable
Author's Description

THUDM: GLM Z1 Rumination 32B is a 32B-parameter deep reasoning model from the GLM-4-Z1 series, optimized for complex, open-ended tasks requiring prolonged deliberation. It builds upon glm-4-32b-0414 with additional reinforcement learning phases and multi-stage alignment strategies, introducing “rumination” capabilities designed to emulate extended cognitive processing. This includes iterative reasoning, multi-hop analysis, and tool-augmented workflows such as search, retrieval, and citation-aware synthesis. The model excels in research-style writing, comparative analysis, and intricate question answering. It supports function calling for search and navigation primitives (`search`, `click`, `open`, `finish`), enabling use in agent-style pipelines. Rumination behavior is governed by multi-turn loops with rule-based reward shaping and delayed decision mechanisms, benchmarked against Deep Research frameworks such as OpenAI’s internal alignment stacks. This variant is suitable for scenarios requiring depth over speed.

Key Specifications
Cost
$$$$
Context
32K
Parameters
32B
Released
Apr 25, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Logit Bias Reasoning Include Reasoning Stop Seed Min P Top P Max Tokens Frequency Penalty Temperature Presence Penalty
Features

This model supports the following features:

Reasoning
Performance Summary

THUDM: GLM Z1 Rumination 32B is a 32B-parameter deep reasoning model designed for complex, open-ended tasks requiring prolonged deliberation. It exhibits strong reliability, with a 91% success rate across benchmarks, indicating consistent and usable responses. However, its "rumination" capabilities, while enhancing depth, result in significantly longer response times, placing it in the 3rd percentile for speed. The model offers competitive pricing, ranking in the 48th percentile. In terms of performance across categories, the model demonstrates a notable strength in General Knowledge, achieving 94.5% accuracy, which is a solid performance. Its Coding and Ethics benchmarks show moderate accuracy at 81.0% for both, though these scores place it in the lower percentiles compared to other models. A significant weakness is observed in Email Classification, where its 88.0% accuracy ranks in the 12th percentile. The model's core strength lies in its ability to perform iterative reasoning and multi-hop analysis, making it suitable for research-style writing and intricate question answering where depth is prioritized over speed. Its function calling capabilities also support agent-style pipelines.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.24
Completion $0.24

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Novita
Novita | thudm/glm-z1-rumination-32b-0414 32K $0.24 / 1M tokens $0.24 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by thudm