GLM 4.7 Flash REAP 23B A3B

Name: GLM 4.7 Flash REAP 23B A3B
Author: Zhipu AI

by Zhipu AIGLM

Open WeightModel weights are publicly available — can be downloaded and self-hosted

Released January 2026203K context≈ 152,064 words23B params

A lean, fast-thinking model that punches above its weight by activating only 3 billion of its 23 billion parameters at inference time — keeping compute costs low without sacrificing too much capability. It handles text tasks with reasonable fluency and follows instructions reliably for its size class. The sparse activation design means it can run efficiently on modest hardware, though complex multi-step reasoning may expose the limits of its active parameter budget.

Capabilities

Capability scores are AI-generated based on model documentation, benchmarks, and technical specifications. Learn more

Reasoning & Logic

Moderate

Instruction Following

GLM 4.7 Flash REAP 23B A3B

by Zhipu AIGLM

Open WeightModel weights are publicly available — can be downloaded and self-hosted

Released January 2026203K context≈ 152,064 words23B params

A lean, fast-thinking model that punches above its weight by activating only 3 billion of its 23 billion parameters at inference time — keeping compute costs low without sacrificing too much capability. It handles text tasks with reasonable fluency and follows instructions reliably for its size class. The sparse activation design means it can run efficiently on modest hardware, though complex multi-step reasoning may expose the limits of its active parameter budget.

Capabilities

Capability scores are AI-generated based on model documentation, benchmarks, and technical specifications. Learn more

Reasoning & Logic

Moderate

Instruction Following

Glossary

InferenceThe process of running a trained model to generate predictions or outputs from new inputs.Inference TimeThe amount of time it takes for a model to process input and generate output after it has been trained.Multi-Step ReasoningThe ability to break down complex problems into smaller steps and solve them sequentially, rather than jumping directly to an answer.Parameter BudgetThe fixed total number of parameters available to allocate across a model's layers and components.ParametersThe learned numerical values in a model — more parameters generally means more capacity but higher compute cost.ReasoningThe model's ability to work through multi-step logical problems and provide justified answers rather than just pattern-matching.Sparse ActivationA technique where only a subset of a model's parameters are used for each input, reducing computational cost while maintaining performance.Thinking ModelA language model trained to generate explicit reasoning steps and internal deliberation before producing a final response, rather than answering immediately.

Capabilities

Capabilities

Use Case Fit

Glossary