Hey PH! 👋 We kept hearing the same question from every revenue leader we talked to: "Which AI model should I actually use for outbound? For account research? For deal prep?" The existing benchmarks test coding, math, and reasoning. None of them test whether an AI can write a cold email your SDR would actually send, or synthesize 8 messy data sources into an account brief without flattening the contradictions. So we built the Revenue AI Index. How it works: Every month we run 8 frontier models (Claude, GPT, Gemini, DeepSeek, Grok, Qwen, Mistral, Llama) through 16 real GTM tasks across 4 categories — Prospecting, Account Intel, Campaigns, and Deal Strategy. Fixed prompts. Blind scoring. 2–3 practitioner reviewers per task. What makes this different from other benchmarks: → Every task is a real revenue workflow, not an academic test → Every prompt is fully inspectable — see exactly what we asked → Every model output is visible — read all 8 and judge for yourself → Scoring is by practitioners, not automated metrics This month's highlights: Claude Opus 4.7 won overall (8.57), but the margins are tight. DeepSeek V4 beat everyone on content repurposing. Gemini 3.1 Pro was surprisingly strong on account intel. Llama 4 Maverick struggled across the board. The rankings shift depending on the workflow. We publish a new edition every month as models update. Would love to hear from this community — which GTM tasks should we benchmark next? What workflows are you testing AI on right now? 🔗 docket.io/resources/tools/revenue-index

Revenue AI Index by Docket

Stop guessing which AI to use for sales & marketing

Stop guessing which AI to use for sales & marketing