Creative writing (Lech Mazur) leaderboard
- GPT-5 leads
- 86.0%
- models measured
- 22
- lead over #2
- +0.4 pts
- median
- 78.5%
Leader on Creative writing (Lech Mazur) · read October 21
On October 21, 2026, 30 of our catalog's models had a Creative writing (Lech Mazur) result, and OpenAI GPT-5 led with 86.0%, ahead of MoonshotAI Kimi K2 0711 at 85.6%.
Short stories that must use a set of required elements, graded by several models. The mean grade out of 10, shown as a share of 10.
Role in the index
Creative writing (Lech Mazur) is shown as published and does not feed any of our indexes. Read the methodology
| # | Model | Result | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|
| 1 | GPT-5OpenAI | 86.0% | 38.8 | $10.00 | |
| 2 | Kimi K2 0711Moonshot AI | 85.6% | Not benchmarked | $2.30 | |
| 3 | Claude Opus 4.1Anthropic | 84.7% | Not benchmarked | $75.00 | |
| 4 | o3 ProOpenAI | 84.4% | Not benchmarked | $80.00 | |
| 5 | o3OpenAI | 83.9% | Not benchmarked | $8.00 | |
| 6 | Gemini 2.5 ProGoogle | 83.8% | 19.1 | $10.00 | |
| 7 | Claude Opus 4RetiredAnthropic | 83.6% | Not benchmarked | $75.00 | |
| 8 | GPT-5 MiniOpenAI | 83.1% | 34.2 | $2.00 | |
| 9 | R1RetiredDeepSeek | 83.0% | Not benchmarked | $2.50 | |
| 10 | Qwen3 235B A22B Thinking 2507Qwen | 82.4% | Not benchmarked | $3.50 | |
| 11 | R1 0528DeepSeek | 81.9% | Not benchmarked | $2.18 | |
| 12 | GPT-4o (2024-11-20)OpenAI | 81.8% | Not benchmarked | $10.00 | |
| 13 | Claude 3.7 SonnetRetiredAnthropic | 81.1% | Not benchmarked | $15.00 | |
| 14 | Claude 3.5 SonnetRetiredAnthropic | 80.3% | Not benchmarked | $15.00 | |
| 15 | Gemma 3 27BGoogle | 79.9% | Not benchmarked | $0.30 | |
| 16 | gpt-oss-120bOpenAI | 77.1% | Not benchmarked | $0.50 | |
| 17 | DeepSeek V3 0324DeepSeek | 77.0% | Not benchmarked | $1.07 | |
| 18 | Grok 3RetiredxAI | 76.4% | Not benchmarked | $15.00 | |
| 19 | Qwen3 30B A3BQwen | 75.3% | Not benchmarked | $0.50 | |
| 20 | o4 MiniOpenAI | 75.0% | Not benchmarked | $4.40 | |
| 21 | Claude 3.5 HaikuRetiredAnthropic | 73.5% | Not benchmarked | $4.00 | |
| 22 | GLM 4.5Z.ai | 73.4% | Not benchmarked | $2.20 | |
| 23 | o1OpenAI | 70.2% | Not benchmarked | $60.00 | |
| 24 | Mistral Large 2407Mistral AI | 69.0% | Not benchmarked | $6.00 | |
| 25 | GPT-4o-mini (2024-07-18)OpenAI | 67.2% | Not benchmarked | $0.60 | |
| 26 | o1-miniRetiredOpenAI | 64.9% | Not benchmarked | $4.40 | |
| 27 | Grok 2RetiredxAI | 63.6% | Not benchmarked | $10.00 | |
| 28 | Phi 4Microsoft | 62.6% | Not benchmarked | $0.14 | |
| 29 | Llama 4 MaverickMeta Llama | 62.0% | Not benchmarked | $1.00 | |
| 30 | o3 MiniOpenAI | 61.7% | Not benchmarked | $4.40 |
Only models in our catalog are listed, and each appears once with its best published run. 215 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out.