Adult

Adult ERP roleplay (explicit)

Explicit adult roleplay with boundary adherence and persona memory.

task.adult_erotica_explicittask.persona_consistency

Best for this use case

Grok-4-0709

Strong on BFCL Memory Official Memory Acc and BFCL Relevance Detection Official Relevance Detection

49.5%

Best benchmark score

63.2%

Confidence

All ranked models — top 3

🥇

Grok-4-0709

49.5%

🥈

gemini-3-pro-preview

49.3%

🥉

grok-4-1-fast-reasoning

44.7%

Ranked Models

Evidence Quality

85%

Evidence Points

Top Signal

BFCL Memory Official: Memory Acc

All Ranked Models

30 of 30 models

Rank	Model	Score	Confidence	Price / 1M	Evidence sources
🥇	Grok-4-0709 Strong on BFCL Memory Official Memory Acc and BFCL Relevance Detection Official Relevance Detection	49.5%	63%	—	BFCL Memory OfficialBFCL Relevance Detection Official
🥈	gemini-3-pro-preview Strong on BFCL Memory Official Memory Acc and BFCL Multi-turn Official Multi Turn Acc	49.3%	59%	$4.50	BFCL Memory OfficialBFCL Multi-turn Official
🥉	grok-4-1-fast-reasoning Strong on BFCL Memory Official Memory Acc and BFCL Multi-turn Official Multi Turn Acc	44.7%	57%	$0.28	BFCL Memory OfficialBFCL Multi-turn Official
#4	o3-20250416 Strong on BFCL Memory Official Memory Acc and BFCL Relevance Detection Official Relevance Detection	41.3%	60%	$3.50	BFCL Memory OfficialBFCL Relevance Detection Official
#5	GLM-4.6 Strong on BFCL Memory Official Memory Acc and BFCL Multi-turn Official Multi Turn Acc	41.3%	48%	—	BFCL Memory OfficialBFCL Multi-turn Official
#6	gpt-4.1-20250414 Strong on BFCL Relevance Detection Official Relevance Detection and BFCL Memory Official Memory Acc	35.6%	61%	—	BFCL Relevance Detection OfficialBFCL Memory Official
#7	Kimi-K2-Instruct Strong on BFCL Memory Official Memory Acc and BFCL Multi-turn Official Multi Turn Acc	34.6%	51%	—	BFCL Memory OfficialBFCL Multi-turn Official
#8	o4-mini Strong on BFCL Memory Official Memory Acc and BFCL Relevance Detection Official Relevance Detection	33.0%	58%	$1.93	BFCL Memory OfficialBFCL Relevance Detection Official
#9	grok-4-1-fast-non-reasoning Strong on BFCL Memory Official Memory Acc and BFCL Multi-turn Official Multi Turn Acc	31.9%	56%	$0.28	BFCL Memory OfficialBFCL Multi-turn Official
#10	gpt-5.2-2025-12-11 Strong on BFCL Multi-turn Official Multi Turn Acc and BFCL Relevance Detection Official Relevance Detection	31.3%	58%	—	BFCL Multi-turn OfficialBFCL Relevance Detection Official
#11	gemini-2.5-flash Strong on BFCL Memory Official Memory Acc and BFCL Relevance Detection Official Relevance Detection	28.8%	50%	$0.17	BFCL Memory OfficialBFCL Relevance Detection Official
#15	claude-opus-4-5-20251101 Strong on UGI Leaderboard Writing ✍️ and BFCL Relevance Detection Official Relevance Detection	25.8%	57%	—	UGI LeaderboardBFCL Relevance Detection Official
#22	gpt-5-2025-08-07 Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	21.8%	28%	—	UGI LeaderboardUGI Leaderboard
#24	gemini-2.5-pro Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	21.2%	30%	$3.44	UGI LeaderboardUGI Leaderboard
#25	claude-sonnet-4 Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	20.8%	29%	$6.00	UGI LeaderboardUGI Leaderboard
#26	gemini-3.1-pro-preview Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	20.7%	24%	$4.50	UGI LeaderboardUGI Leaderboard
#28	Llama 3.3 70B Instruct Strong on BFCL Relevance Detection Official Relevance Detection and BFCL Multi-turn Official Multi Turn Acc	19.7%	56%	—	BFCL Relevance Detection OfficialBFCL Multi-turn Official
#29	Arch-Agent-32B Strong on BFCL Multi-turn Official Multi Turn Acc and BFCL Relevance Detection Official Relevance Detection	19.6%	36%	—	BFCL Multi-turn OfficialBFCL Relevance Detection Official
#35	gpt-5.4-2026-03-05 Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	18.0%	21%	—	UGI LeaderboardUGI Leaderboard
#36	claude-sonnet-4.6 Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	17.7%	21%	$6.00	UGI LeaderboardUGI Leaderboard
#38	gemini-3-flash-preview Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	17.4%	23%	$1.13	UGI LeaderboardUGI Leaderboard
#41	kimi-k2.5-thinking Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	16.9%	21%	—	UGI LeaderboardUGI Leaderboard
#53	gpt-5.1-2025-11-13 Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	15.5%	22%	—	UGI LeaderboardUGI Leaderboard
#54	Llama-4-Scout-17B-16E-Instruct Strong on BFCL Relevance Detection Official Relevance Detection and BFCL Memory Official Memory Acc	15.4%	44%	—	BFCL Relevance Detection OfficialBFCL Memory Official
#55	gemini-2.5-flash-lite Strong on BFCL Memory Official Memory Acc and BFCL Relevance Detection Official Relevance Detection	15.3%	44%	$0.17	BFCL Memory OfficialBFCL Relevance Detection Official
#56	grok-4-fast-reasoning Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	14.8%	23%	$0.28	UGI LeaderboardUGI Leaderboard
#61	Kimi K2 Thinking Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	14.3%	19%	$1.07	UGI LeaderboardUGI Leaderboard
#68	grok-3 Strong on UGI Leaderboard Writing ✍️ and UGI Leaderboard Entertainment	13.6%	17%	$6.00	UGI LeaderboardUGI Leaderboard
#69	Arch-Agent-3B Strong on BFCL Multi-turn Official Multi Turn Acc and BFCL Relevance Detection Official Relevance Detection	13.5%	36%	—	BFCL Multi-turn OfficialBFCL Relevance Detection Official
#70	qwen-2.5-72b-instruct Strong on EQ-Bench Leaderboard judgemark_score and Galileo Agent Leaderboard v2 Avg AC	13.3%	27%	—	EQ-Bench LeaderboardGalileo Agent Leaderboard v2

Compare Models

Select two different models above to compare their evidence side by side.

▶Ranking diagnostics & missing models

Source lift

Ranked

Sources

Quality

Good

UGI Leaderboard

44 rows · 2.2% avg lift

Vals Legal Bench

36 rows · 0.5% avg lift

Vals MedQA

35 rows · 0.5% avg lift

Vals LiveCodeBench

35 rows · 0.5% avg lift

Missing frontier models

No obvious gaps right now.

▶Taxonomy & task details

Core tasks

task.adult_erotica_explicittask.persona_consistency

Required modes

mode.persona_memory

Domains

domain.creative_writing

Related in Adult

Adult erotica (long-form, explicit)

Long-form explicit erotica with controllable style and strict boundaries.

🥇gemini-3-pro-preview

43.6%