Traditional grading2–3 days
GradeLogic first-pass2–5 hours
Estimated time savedup to 70–85%
Ready to grade your first exam?
Upload your first assessment to start seeing insights and saving time.
Run on any LLM — cloud or local
GradeLogic decouples grading logic from specific models. Assign different LLMs to different stages of the grading pipeline based on cost, context window, or privacy requirements.
- Cloud: GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro
- Local (Ollama): Llama 3, Mistral, Command R+
Model Assignments
| Pipeline Step | Assigned Model |
|---|---|
| Upload | GPT-4o Vision |
| Matching | Claude Haiku |
| Grading | Gemini 1.5 Pro |
Zero infrastructure. Zero lock-in.
GradeLogic operates entirely in your browser. It communicates directly with your chosen LLM APIs and stores all data locally. There is no central server, meaning your grading data remains entirely in your control.
No server No cloud No account
Browser
LLM API
Local Storage
Connect any AI model — internal or external
GradeLogic decouples grading logic from the model — wire in OpenAI, Anthropic, Google, or Ollama without rewriting your workflow.
See supported providers O
OpenAI Connected
A
Anthropic Configure
G
Gemini Configure
OL
Ollama Connected
AZ
Azure OpenAI Configure
+
Custom API Configure
Your data stays on your device, always
- No server infrastructure to maintain
- No cloud storage or sync
- No account required to use
Your Browser
LLM API
PII never leaves your institution