NC Bench
A comprehensive benchmark for creative writing models.
Created by Novelcrafter.
This project is in early access and still work in progress.
Focused on Creativity
NC Bench benchmarks creativity-focused tasks in LLMs, like creative writing, instruction following, utility, tooling, and language skills.
AI as an Assistant
We test how well models enhance the writing process through text manipulation, idea generation, summarization, and translation.
Comprehensive Testing
From generating quality prose to hallucination-free extraction, NC Bench evaluates capabilities using objective tests, not AI judges.
Benchmark Overview
Tests
14Scenarios
112Models
147Results
140,411Samples per Run
8.21 runs/scen.Category Distribution
Shows the number of scenarios in each category. Some scenarios may be in multiple categories.
Top Results
Top Models
| 95.02% | Claude Opus 4.6 (Reasoning) |
| 94.54% | Qwen3.6 Max Preview |
| 94.37% | Gemini 3.1 Pro (Preview) |
Tooling
| 100.00% | Claude Opus 4.6 (Reasoning) |
| 100.00% | Qwen3.6 Max Preview |
| 100.00% | Z.AI GLM 5.1 |
Creative Writing
| 91.17% | GPT-5.4 (Reasoning) |
| 90.94% | GPT-5.4 |
| 90.51% | GPT-5.4 (Reasoning, Low) |
Language
| 100.00% | Qwen3.6 Max Preview |
| 100.00% | Claude Sonnet 4.6 |
| 100.00% | DeepSeek-V2 Chat |
Utility
| 99.91% | Gemini 3.1 Pro (Preview) |
| 98.93% | Claude Opus 4.6 (Reasoning) |
| 98.67% | o4 Mini High |
Text Editing
| 99.13% | Claude Sonnet 4 |
| 99.02% | Claude Sonnet 4.5 |
| 98.90% | Z.AI GLM 5.1 |
Hallucination
| 99.93% | Stealth: Aurora Alpha |
| 99.69% | Nemotron 3 Super |
| 99.06% | o4 Mini High |