wesleysimplicio/Simplicio-27B

VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Oct 4, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

Simplicio 27B, developed by Wesley Simplicio at SimpleTI, is a 27.36 billion parameter LoRA fine-tune of Qwen3.8-27B, designed for code change requests. This model specializes in generating SEARCH/REPLACE patches rather than entire files, operating within a 32,768-token context window. It processes requests through five tagged phases, focusing on precise, line-level code modifications for development workflows.

Loading preview...

Simplicio 27B: Code Change Agent with SEARCH/REPLACE Patches

Simplicio 27B is a 27.36 billion parameter LoRA fine-tune of the Qwen3.8-27B model, developed by Wesley Simplicio at SimpleTI. Its primary innovation lies in its ability to respond to code change requests by generating precise SEARCH/REPLACE patches, rather than outputting entire modified files. This approach aims to streamline code modification workflows by focusing on only the lines that require changes.

Key Capabilities

  • Patch-based Code Changes: Generates code modifications using a SEARCH/REPLACE format, targeting specific lines of code.
  • Phased Response Generation: Structures its output into five distinct, tagged phases: <orient>, <plan>, <patch>, <validate>, and <deliver>, guiding the user through the change process.
  • Code Agent Focus: Designed specifically for automated code modification tasks, aiming for surgical precision in diffs.
  • Qwen3.8-27B Base: Built upon a robust base model with a 32,768-token context length, leveraging its architectural strengths.
  • Research Release: Currently a research release, demonstrating a 46.7% pass rate on its internal held-out set of 120 tasks.

Good For

  • Automated Code Refactoring: Ideal for scenarios requiring automated, precise modifications to existing codebases.
  • Developer Tool Integration: Can be integrated into development environments or CI/CD pipelines to automate small, targeted code changes.
  • Experimentation with Code Agents: Useful for researchers and developers exploring advanced code agent capabilities and patch-based generation.
  • Specific Code Fixes: Suited for tasks where changes are localized and can be expressed as SEARCH/REPLACE operations.

Finetunes
1 models