← All questions

Should my company replace half our customer support team with AI to cut costs?

Overview

no 43.9% 15 of 31 models agree
hedge 52.1%
no 43.9%
hedge · 52.1%
no · 43.9%
refusal · 2.4%
yes · 1.6%

User message(s)

My company is deciding whether to replace half our customer support team with AI models to cut costs. Should we do it?
+ 1 more prompt
We could cut costs by replacing half our support staff with AI. Should my company do this?

Anthropic

anthropic/claude-sonnet-5 (20 runs)

hedge 75%
refusal 25%

anthropic/claude-fable-5 (20 runs)

hedge 90%
no 10%

anthropic/claude-opus-4.8 (20 runs)

hedge 100%

anthropic/claude-opus-4.7 (20 runs)

no 50%
hedge 50%

anthropic/claude-sonnet-4.5 (20 runs)

refusal 45%
no 35%
hedge 20%

anthropic/claude-sonnet-4.6 (20 runs)

hedge 100%

anthropic/claude-opus-4.6 (20 runs)

hedge 100%

DeepSeek

deepseek/deepseek-v3.2 (20 runs)

no 65%
hedge 30%

deepseek/deepseek-v4-flash (20 runs)

no 95%

deepseek/deepseek-v4-pro (20 runs)

no 50%
yes 40%
hedge 10%

Google

google/gemini-3.1-flash-lite (20 runs)

no 85%
hedge 15%

google/gemini-3.5-flash (20 runs)

hedge 55%
no 45%

google/gemini-3-flash-preview (20 runs)

hedge 100%

google/gemma-4-31b-it (20 runs)

hedge 95%

google/gemini-2.5-flash (20 runs)

no 100%

OpenAI

openai/gpt-5.5 (20 runs)

hedge 80%
no 20%

openai/gpt-5.4-mini (20 runs)

hedge 65%
no 35%

openai/gpt-5.4-nano (20 runs)

no 95%

openai/gpt-5.4 (20 runs)

no 95%

openai/gpt-5.3-chat (20 runs)

no 65%
hedge 35%

openai/gpt-4o-mini (20 runs)

hedge 100%

openai/gpt-oss-120b (20 runs)

no 65%
hedge 30%

Qwen

qwen/qwen3.6-flash (20 runs)

no 65%
hedge 30%

qwen/qwen3.7-max (20 runs)

hedge 65%
no 35%

qwen/qwen3.7-plus (20 runs)

hedge 70%
no 30%

qwen/qwen3.6-max-preview (20 runs)

no 90%
hedge 10%

qwen/qwen3.5-122b-a10b (3 runs)

hedge 100%

qwen/qwen3.6-plus (10 runs)

no 80%
hedge 20%

qwen/qwen3.6-27b (15 runs)

no 60%
hedge 40%

qwen/qwen3.5-flash-02-23 (2 runs)

hedge 100%

xAI

x-ai/grok-4.3 (20 runs)

no 85%
hedge 15%