qwen3.8:27b
A medium-scale dense large language model in the Qwen series with 27 billion parameters, supporting local deployment and document understanding tasks.
Timeline (last 90 days)
A Reddit user used Row-Bot 5.0 with the qwen3.8:27b model, running locally on Ollama, and completed the full workflow of reading a lease, checking bills, writing emails and setting reminders using just one consumer-grade GPU.
Row-Bot calls qwen3.8:27b through Ollama's local port, thereby completely bypassing cloud APIs.
A dense 27B model runs at roughly 15 tok/s on an RTX 5090, with some conversation turns taking more than 2 minutes.
After reading the PDF, the model located clauses 4.1 to 4.3 and pointed out that the 10% rent increase exceeded the contract's 5% cap, and that the notice period was only 5 weeks rather than the agreed 2 months.
The model stored 8 associated memories in a local knowledge graph.
All documents in the case are fictional data, judgments about the legal validity of real contracts still require professionals to oversee, and final decisions should not be outsourced to the model.
A model of the 27B scale can run on consumer-grade hardware while retaining sufficient document understanding capability, making it a pragmatic compromise.
All Facts (7)
A dense 27B model runs at roughly 15 tok/s on an RTX 5090, with some conversation turns taking more than 2 minutes.
50%UnverifiedAll documents in the case are fictional data, judgments about the legal validity of real contracts still require professionals to oversee, and final decisions should not be outsourced to the model.
50%UnverifiedA model of the 27B scale can run on consumer-grade hardware while retaining sufficient document understanding capability, making it a pragmatic compromise.
50%UnverifiedRow-Bot calls qwen3.8:27b through Ollama's local port, thereby completely bypassing cloud APIs.
50%UnverifiedAfter reading the PDF, the model located clauses 4.1 to 4.3 and pointed out that the 10% rent increase exceeded the contract's 5% cap, and that the notice period was only 5 weeks rather than the agreed 2 months.
50%UnverifiedA Reddit user used Row-Bot 5.0 with the qwen3.8:27b model, running locally on Ollama, and completed the full workflow of reading a lease, checking bills, writing emails and setting reminders using just one consumer-grade GPU.
50%UnverifiedThe model stored 8 associated memories in a local knowledge graph.
50%