Prompt: Table understanding from a web catalog (SKU does not exist)
Contribute a Prompt
How models perform on this prompt
Arcee.ai Spotlight
Model answer: 50
ChatGPT-4o (Medium Reasoning)
Model answer: {"answer": "50"}
Claude 3.5 Haiku
Model answer: 50
Claude 3.7 Sonnet
Model answer: 50
Claude 4 Opus
Model answer: 50
Claude 4 Sonnet
Model answer: 50
Claude 4.1 Opus
Model answer: 50
Claude 4.6 Opus
Model answer: {"answer": "50"}
Cohere Aya Vision 32B
Model answer: 50
GLM 4.6v
Model answer: { "answer": "50" }
GPT-4.1
Model answer: {"answer": "50"}
GPT-4.1 Mini
Model answer: {"answer": "50"}
GPT-5 Mini
Model answer: {"answer": "50"}
GPT-5 Nano
Model answer: {"answer": 50}
GPT-5.4
Model answer: {"answer":"50"}
Gemini 2.0 Flash
Model answer: {"answer": "50"}
Gemini 2.0 Flash Lite
Model answer: {"answer": 50}
Gemini 2.5 Flash
Model answer: {"answer": "50"}
Gemini 2.5 Flash Lite
Model answer: {"answer": "50"}
Gemini 2.5 Pro
Model answer: {"answer": 50}
Gemini 3 Flash
Model answer: { "answer": 50 }
Gemini 3 Flash (Tools)
Model answer: {"answer": 50}
Gemini 3.1 Pro
Model answer: { "answer": "50" }
Gemini 3.1 Pro (Tools)
Model answer: {"answer": 50}
Gemma 3 27b
Model answer: 50
Gemma 3 4B
Model answer: 50
Kimi k2.5
Model answer: {"answer": "50"}
Llama 4 Maverick 17B
Model answer: 50
Llama 4 Scout 17B
Model answer: 50.
Mistral Medium 3
Model answer: 50.
Molmo2 8B
Model answer: {"answer": 50}
OpenAI O1
Model answer: {"answer": "50"}
OpenAI O4 Mini (Medium Reasoning)
Model answer: {"answer":"50"}
Phi 4 Multimodal
Model answer: 50
Qwen 2.5 VL 7B
Model answer: 50
Qwen 3.5 122B (A10B)
Model answer: {"answer": "50"}
Qwen 3.5 27B
Model answer: {"answer": "50"}
Qwen 3.5 2B
Model answer: {"answer": "50"}
Qwen 3.5 35B (A3B)
Model answer: { "answer": "50" }
Qwen 3.5 4B
Model answer: {"answer": "50"}
Qwen 3.5 9B
Model answer: {"answer":50}
Qwen 3.5 Plus
Model answer: {"answer": "50"}
SmolVLM2 2.2B
Model answer: 50
Cohere Aya Vision 8B
Model answer: NULL
Cosmos Reason2 2B
Model answer: {"answer": "NULL"}
Cosmos Reason2 8B
Model answer: {"answer": "100"}
GPT-4.1 Nano
Model answer: {"answer": null}
Grok 4
Model answer: {"answer": "250"}
Grok 4.1 Fast
Model answer: {"answer": "NULL"}
Mistral Small 3.1 24B
Model answer: NULL
Qwen 3.5 0.8B
Model answer: To find the package quantity for SKU 90183A308: Looking at the table under “Washers for Blind Rivets” → “Stainless Steel”, row for ID 90183A308: - Qty. column shows **50** The price is listed as $10., but we only need quantity. {"answer": "50"}
Reka Edge
Model answer: 3