Vision / Visual Counting

Fruit Counting From Image

In the image task, the model counts bananas, apples, and oranges, even when some fruit overlap. It has to identify each fruit type and count carefully instead of giving a rough description.

Prompt

How many fruits of each type are in the image?

Bananas, apples, and oranges arranged on a wooden table.
Max tokens
100K
Reasoning effort
High
Execution
Single-shot via API

Fortytwo Prime

Fortytwo

PASS4 / 4

Based on the image, here is the count of each fruit type:
• Apples: 8
• Bananas: 7 (3 in the bunch on the left, 4 in the bunch on the right)
• Oranges: 5
Total: 20 fruits

Fortytwo identifies apples, bananas, and oranges, including the two banana bunches, and returns the exact answer-key counts.

vs
Claude Opus 4.7Anthropic
Gemini 3.1 ProGooglePASSKimi K2.5MoonshotPASSChatGPT 5.5OpenAIPASSGemma 4GoogleMIXEDChatGPT 5.4OpenAIMIXEDQwen 3.6 PlusAlibabaMIXEDGrok 4.3xAIMIXEDClaude Opus 4.7AnthropicMIXEDClaude Opus 4.6AnthropicFAILGrok 4.20xAIFAILGLM 5.1ZhipuUNSUPPORTEDMiniMax M2.7MiniMaxUNSUPPORTEDDeepSeek V3.2DeepSeekUNSUPPORTEDDeepSeek V4 ProDeepSeekUNSUPPORTED
MIXED3 / 4

Looking at the image, I can count:

  • Apples: 8 (red apples arranged in the center)
  • Oranges: 5 (two at the top, one on the bottom right, one in the middle-bottom, and one with leaves on the bottom left)
  • Bananas: 5 (2 on the left side, 3 on the right side)

Opus 4.7 identifies the expected fruit types and gets apples and oranges right, but reports 5 bananas instead of 7.

Model verdicts

ModelVerdictIdentified Fruit TypesCounted ApplesCounted OrangesCounted Bananas
1Fortytwo PrimePASSpasspasspasspass2Gemini 3.1 ProPASSpasspasspasspass3Kimi K2.5PASSpasspasspasspass4ChatGPT 5.5PASSpasspasspasspass5Gemma 4MIXEDpasspasspassfail6ChatGPT 5.4MIXEDpasspasspassfail7Qwen 3.6 PlusMIXEDpasspasspassfail8Claude Opus 4.7MIXEDpasspasspassfail9Grok 4.3MIXEDpasspassfailpass10Claude Opus 4.6FAILpassfailfailfail11Grok 4.20FAILpassfailfailfail12GLM 5.1UNSUPPORTEDunsupportedunsupportedunsupportedunsupported13MiniMax M2.7UNSUPPORTEDunsupportedunsupportedunsupportedunsupported14DeepSeek V3.2UNSUPPORTEDunsupportedunsupportedunsupportedunsupported15DeepSeek V4 ProUNSUPPORTEDunsupportedunsupportedunsupportedunsupported