List your company

Question Answering on SimpleQA Verified

82.5AccuracyHEART
0.44421.74743.0564.353Sep 26, 2025Nov 9, 2025Dec 24, 2025Feb 7, 2026Mar 24, 2026May 8, 2026Jun 22, 2026

Evaluation results

MethodLinks
HEARTModel=Gemini 3 Pro2025.0982.5
Self ReflectionModel=Gemini 3 Pro2025.0982.4
CoTModel=Gemini 3 Pro2025.0981.38
WaitModel=Gemini 3 Pro2025.0980.18
Self ReflectionModel=GPT-5-nano2025.0978.39
HEARTModel=GPT-5-nano2025.0976.84
WaitModel=GPT-5-nano2025.0976.28
CoTModel=GPT-5-nano2025.0974.58
VanillaModel=Gemini 3 Pro2025.0972.27
VanillaModel=GPT-5-nano2025.0970.83
HEARTModel=Deepseek-Reasoner2025.0968.96
WaitModel=Deepseek-Reasoner2025.0964.39
HEARTModel=Gemini 3 Flash2025.0964.29
CoTModel=Gemini 3 Flash2025.0963.85
Self ReflectionModel=Gemini 3 Flash2025.0961.39
WaitModel=Gemini 3 Flash2025.0961.28
CoTModel=Deepseek-Reasoner2025.0961.2
HEARTModel=Claude 4 Sonnet2025.0959.39
Self ReflectionModel=Deepseek-Reasoner2025.0958.84
Self ReflectionModel=Claude 4 Sonnet2025.0958.29
HEARTModel=Gemini 2.5 Pro2025.0957.29
CoTModel=Claude 4 Sonnet2025.0956.58
WaitModel=Gemini 2.5 Pro2025.0955.75
WaitModel=Claude 4 Sonnet2025.0955.29
Self ReflectionModel=Gemini 2.5 Pro2025.0952.38
CoTModel=Gemini 2.5 Pro2025.0951.19
HEARTModel=Gemini-2.5-Flash (Thinking Budget set to 0)2025.0949.88
HEARTModel=Claude-4-Sonnet (Thinking Off)2025.0948.97
HEARTModel=Deepseek-Chat2025.0948.14
HEARTModel=Gemini 2.5 Flash2025.0947.38
Self ReflectionModel=Gemini-2.5-Flash (Thinking Budget set to 0)2025.0946.3
WaitModel=Gemini-2.5-Flash (Thinking Budget set to 0)2025.0946.2
WaitModel=Gemini 2.5 Flash2025.0945.84
CoTModel=Gemini-2.5-Flash (Thinking Budget set to 0)2025.0945.4
WaitModel=Claude-4-Sonnet (Thinking Off)2025.0945.28
CoTModel=Deepseek-Chat2025.0945.23
CoTModel=Claude-4-Sonnet (Thinking Off)2025.0944.39
Self ReflectionModel=Deepseek-Chat2025.0944.29
CoTModel=Gemini 2.5 Flash2025.0944.2
WaitModel=Deepseek-Chat2025.0944.19
Self ReflectionModel=Gemini 2.5 Flash2025.0942.18
Self ReflectionModel=Claude-4-Sonnet (Thinking Off)2025.0941.49
APEXModel=Gemini 2.5 Flash2026.0635
VanillaModel=Gemini 3 Flash2025.0932.77
VanillaModel=Gemini 2.5 Pro2025.0932.7
VanillaModel=Deepseek-Reasoner2025.0931.92
VanillaModel=Gemini-2.5-Flash (Thinking Budget set to 0)2025.0929.5
VanillaModel=Deepseek-Chat2025.0929.38
GEPAModel=Gemini 2.5 Flash2026.0628.8
VanillaModel=Gemini 2.5 Flash2025.0928.48
APO (|D|)Model=Gemini 2.5 Flash2026.0627.4
APO (|D|/2)Model=Gemini 2.5 Flash2026.0625
VanillaModel=Claude 4 Sonnet2025.0924.98
VanillaModel=Claude-4-Sonnet (Thinking Off)2025.0923.75
Initial PromptModel=Gemini 2.5 Flash2026.0623.6
HEARTModel=Gemma3 12b Instruct2025.0918.47
HEARTModel=Gemma3 4b Instruct2025.0915.61
WaitModel=Gemma3 12b Instruct2025.0914.29
CoTModel=Gemma3 12b Instruct2025.0912.38
APO (|D|)Model=Gemma 3 27B2026.0611.5
APEXModel=Gemma 3 27B2026.0611.5
Self ReflectionModel=Gemma3 12b Instruct2025.0910.29
Self ReflectionModel=Gemma3 4b Instruct2025.0910
Initial PromptModel=Gemma 3 27B2026.069.4
GEPAModel=Gemma 3 27B2026.069.4
WaitModel=Gemma3 4b Instruct2025.099.04
APO (|D|/2)Model=Gemma 3 27B2026.068.4
Diff.Backbone=Qwen3-4B-Base, Synthesis Strategy=Difficulty-weighted2026.068.2
KDoSBackbone=Qwen3-4B-Base2026.068
Uni.Backbone=Qwen3-4B-Base, Synthesis Strategy=Uniform2026.067.9
CoTModel=Gemma3 4b Instruct2025.097.89
SPBackbone=Qwen3-4B-Base, Training Data=Seed Pool2026.066.8
Qual.Backbone=Qwen3-4B-Base, Synthesis Strategy=Quality-filtered2026.066.8
Rand.Backbone=Qwen3-4B-Base, Synthesis Strategy=Random2026.066.7
VanillaModel=Gemma3 12b Instruct2025.095.83
VanillaModel=Gemma3 4b Instruct2025.094.7
BaseBackbone=Qwen3-4B-Base2026.063.6