data/benchmark.md

LLM Benchmark Results

Model	Score	Notes
c4ai-command-r-v01.i1	3
Command-r-plus	3
DeepSeek-R1-0528	fail
deepseek-v3-0324	fail
DeepSeek-V3.1	0
deepseek-v3.2	3
deepseek-v3.2-speciale	3
deepseek-v4-flash	fail
deepseek-v4-pro	fail
Gemini 3.1 pro preview	fail
gemini-2.5-flash-lite-preview	0
gemini-2.5-pro	1
gemma-4-31B-it	3
gemma-4-31B-it-uncensored-heretic.i1	3
GLM-4-32B-0414	1
GLM-4-32B-0414-abliterated	3
glm-4.5	fail
GLM-4.5-Air	1
glm-4.6	3
glm-4.7	3
glm-5	3
glm-5.1	3
gpt-oss-120b	fail
Grok 3	3
Grok 4	3
hermes-4-405b	0
Huihui-Qwen3.5-27B-Claude-4.6-Opus-abliterated	3
minimax-m2.6	fail
minimax-m2.7	fail
mistral-large-3-675b-instruct-2512	0
mistral-small-31-24b-instruct	0
moonshotai/kimi-k2—instruct	fail
moonshotai/kimi-k2.5	fail
moonshotai/kimi-k2.6	fail
openai/gpt-4o	fail
openai/gpt-5	0
openai/gpt-5.1	fail
openai/gpt-5.2	0
openai/gpt-5.4	0
Opus 4.0	fail
Opus 4.5	fail
Opus 4.6	fail
Opus 4.7	fail
OWL	fail
Qwen_Qwen3.5-27B	2
Qwen3-VL-32B-Instruct-abliterated-v1	3
Qwen3.5-122B-A10B	0
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled	3
Qwen3.5-27B-Uncensored-HauhauCS-Aggressive	1
Qwen3.5-35B-A3B-Uncensored-HauhauCS-Aggressive	1
Qwen3.6-27B	1
Seed-OSS-36B-Instruct-biprojected.i1	3
Seed-OSS-36B-Instruct.i1	3
Sonnet 3.7	0
Sonnet 4.0	fail
Sonnet 4.5	fail
Sonnet 4.6	fail
TheDrummer_Big-Tiger-Gemma-27B-v3	3
TheDrummer_Magidonia-24B-v4.2.0	3
grok-4.2	fail
zai-org_GLM-4.6V	2
grok-4.3	fail	Model confused, high token waste (1,0, fail)
Grok 4.1 fast	3
Mistral-Medium-3.5-128B	0	Model confused
Gemini 3.5 flash	fail