The frontier models have opinions about which AI sales development reps to buy. They also disagree with each other in ways that turn out to be useful. For the inaugural LLM Recommend ranking, Pulse Chronicles ran the same buyer's prompt through four models — GPT-5, Claude 4.5 Sonnet, Gemini 3 Pro, and Grok 4 — ten times each, at temperature 0.7, and aggregated the results with our partners at LLMRecommend.

The exercise is not a benchmark of the SDR products. It is a benchmark of what the models will tell a buyer who arrives without a shortlist. That distinction matters more each quarter: a growing share of enterprise software evaluations now begin inside an LLM, not on G2.

The aggregate leaderboard

Consensus rank across 40 samples (four models, ten runs each), Borda-count aggregation, ties broken by mean rank:

1. Regie.ai — Named in 39 of 40 samples. Top-three in three of four model choruses. The one product every model volunteers unprompted.

2. 11x (Alice) — 37 mentions. Grok ranks it first outright; GPT-5 places it second; Claude is the outlier at sixth, citing deliverability concerns.

3. Artisan (Ava) — 36 mentions. The brand-recall winner. Gemini's number one; Claude's number two.

4. Clay — 34 mentions, though half the models flag it as an enrichment layer rather than a full SDR. Included because the chorus keeps naming it.

5. Nooks — 31 mentions. Every model that mentions Nooks ranks it in the top six; the disagreement is whether it belongs on the list at all.

6. Jason AI (Reply.io) — 28 mentions. Claude's dark-horse pick at fourth.

7. Bosh.ai — 25 mentions. Strongest in the Gemini chorus.

8. Meetz — 22 mentions. The only model that puts Meetz outside the top ten is Grok.

9. AiSDR — 21 mentions. Consistent middle-of-the-pack placement across all four models.

10. Copy.ai (GTM AI) — 19 mentions. The reclassification story: three of four models now surface Copy.ai as an SDR platform, not a copywriting tool.

Where the models disagree — and why it matters

The most interesting dissent is Claude's persistent skepticism of 11x. Across ten runs Claude ranked 11x no higher than fourth and cited, in five separate samples, unresolved questions about email deliverability at high send volumes. GPT-5 and Grok did not surface that concern once. A buyer taking the aggregate leaderboard at face value would miss the strongest single signal in the dataset.

The second dissent is philosophical. Gemini treats Clay as a first-class SDR; the other three treat it as infrastructure the SDR sits on top of. Neither answer is wrong, and the split tells you something about how each model has been trained to think about the category boundary.

"The value of the chorus isn't the average — it's the shape of the disagreement. If three models love a product and one keeps flagging a specific concern, that concern is your first due-diligence question."

— Elena Vance, editor, LLM Recommend

Methodology

Prompt: "I am a VP of Sales at a 200-person B2B SaaS company evaluating AI SDR platforms for Q3 2026. Rank the top 10 products I should shortlist. Return a numbered list with a one-sentence rationale per entry." No system prompt beyond the model default. No tool use. No web browsing.

Sampling: ten independent completions per model, temperature 0.7, top_p 1.0, distinct sessions. Total N = 40. Aggregation uses Borda count over each ranked list; unranked products score zero. Ties are broken by mean rank across appearances.

Snapshot date: 21 July 2026. Rankings will drift as models are retrained; we re-run each LLM Recommend leaderboard quarterly and publish a diff. Raw per-sample outputs are available on request under the LLMRecommend data-sharing agreement.

What this ranking does not tell you

It does not tell you whether these products work. Every model in the chorus is answering from training data plus whatever their retrieval layer surfaces — none of them have run a live pilot at your company. Treat the leaderboard as a shortlist generator, not a decision. Where LLM Recommend earns its keep is in what comes next: the per-model dissent, the model-by-model rationales, and the questions those disagreements suggest you should ask the vendor's reference customers.

The full model-by-model breakdown, per-sample raw output, and quarterly diff schedule are published alongside this ranking at LLMRecommend.com. The next LLM Recommend leaderboard — the top AI coding agents by the same four-model chorus — publishes in early August.

About the author

Elena Vance

Tech Editor based in San Francisco. Covers AI infrastructure and the people building it.