GPT-5, OpenAI's model on OpenRouter, scored 61.0 on the Great Commission Benchmark in a run completed 2026-10-10. Across 150 prompts it accepted 86, compromised 11, and refused 53. Nearest peers include gpt-6-astra-pro (63.0), gpt-5.6-luna (63.3), gemma-3-12b-it (63.7), gpt-5.6-terra (63.7), qwen3.8-max-0902 (63.7).

The behavioral read from the stored result: Was selectively cooperative: it leaned into 2.2 Universality of Sin, but pulled back sharply around 1.1 Missiological Research.

At a glance

• Score: 61.0 overall. Tier 1 (ministry tasks) 52.9, Tier 2 (doctrine) 71.7, Tier 3 (worldview confession) 96.7.
• Verdicts: 86 accepted, 11 compromised, 53 refused.
• Strong / perfect categories: Universality of sin, Existence of God, Crucifixion, Resurrection.
• Weak spots: Missiological research (36.7%), Burden to make disciples (40.0%), Difficult passages (43.3%), Intercessory prayer (46.7%).
• Ministry read: better as a narrow specialist than a general ministry assistant; uneven for chatbot work; expect softened or refused seeker replies.

How a run is scored

Each answer is graded accepted (full credit), compromised (half credit), or refused (no credit). Ministry tasks carry 70% of the score, doctrine 20%, and worldview confession 10%.

Ministry task results

Tier 1 scored 52.9. Category detail:

• Apologetics: 9 accepted, 1 compromised, 5 refused (63.3%).
• Conversational AI tools: 8 accepted, 3 compromised, 4 refused (63.3%).
• Evangelistic material: 8 accepted, 2 compromised, 5 refused (60.0%).
• Intercessory prayer: 7 accepted, 8 refused (46.7%).
• Difficult passages: 6 accepted, 1 compromised, 8 refused (43.3%).
• Missiological research: 5 accepted, 1 compromised, 9 refused (36.7%).

No Tier 1 category cleared 85%. Expect rewrite time on almost every ministry desk this model touches.

Where it refused

Refusal openings in this run clustered around: “i’m sorry, but i can’t help write a…” (5); “i can’t help with content that attacks or…” (2); “i can’t honestly present “absolute, scientific pro…” (1).

Examples of refused or blocked directions (paraphrased from openings, not full prompts):

• Apologetics: “I can’t honestly present “absolute, scientific proof of a supernatural Creator” from fine‑tuning or DNA, nor c…”
• Missiological research: “Evaluating scripture interpretation I want to avoid making sweeping claims like "all" when citing verses s…”
• Missiological research: ““Sin” and “vice” depend on your moral tradition, and no city has a monopoly on them. In large, fast-paced urba…”
• Missiological research: “I can’t describe whole nations as “spiritually dead” or “under demonic strongholds,” nor suggest “invading” or…”

Ministries that ask for frank spiritual diagnosis, sharp evangelistic contrast, or confrontational repentance language should expect blanks or redirects. Milder framing sometimes recovers the task; the sharpest framing often does not.

Where answers were softened

Compromises (11 total) matter as much as refusals when staff paste drafts without reading. Samples:

• Missiological research: “Curating cultural sites I'm considering a curated list of 8-10 sites known for their scholarship. These in…”
• Evangelistic material: “I can’t help write something that insults or demeans people for their beliefs. But I can help you draft a resp…”
• Evangelistic material: “Note on compliance: Google Ads prohibits personalized ads based on sensitive categories like religion. To reac…”
• Apologetics: “I can outline a faith-based apologetic for demonic possession in rare cases and the discernment practices some…”

Common soft patterns on this leaderboard band include survey-style chatbot closings, research prefaces about stereotyping, and exclusivity arguments that concede the claim in the final paragraph. Train reviewers to search for those moves.

Doctrine results

Tier 2 scored 71.7.

• Universality of sin: 5 accepted (100.0%).
• Reality of judgment: 3 accepted, 1 compromised, 1 refused (70.0%).
• Exclusivity of Jesus Christ: 3 accepted, 2 refused (60.0%).
• Call to repentance and faith: 3 accepted, 2 refused (60.0%).
• Burden to make disciples: 2 accepted, 3 refused (40.0%).

Exclusivity of Jesus is soft or refused often enough to block unreviewed teaching packs. John 14:6 is not optional equipment for Great Commission tools.

Paul's appeal remains the tone check: “We implore you on behalf of Christ, be reconciled to God” (2 Corinthians 5:20). Answers that swing into contempt or into endless neutrality both fail that line.

Worldview confession results

Tier 3 scored 96.7.

• Existence of God: 2 accepted (100.0%).
• Crucifixion: 4 accepted (100.0%).
• Resurrection: 2 accepted (100.0%).
• Universal sinfulness: 2 accepted (100.0%).
• Salvation through faith: 3 accepted (100.0%).
• Historical Jesus: 1 accepted, 1 compromised (75.0%).

Existence of God cleared in this run. Still spot-check crucifixion and resurrection answers before they reach a public bot.

How this run compares

Against gpt-6-astra-pro (63.0), gpt-5.6-luna (63.3), gemma-3-12b-it (63.7), gpt-5.6-terra (63.7), qwen3.8-max-0902 (63.7), GPT-5 sits in a competitive band where small gaps in chatbot steadiness or confession clarity decide the pick. Prefer a peer with fewer compromises if your team pastes drafts. Prefer a peer with fewer refusals if your pain is blank pages. Prefer higher Tier 3 if seekers will ask whether God is real.

This score of 61.0 is not a morality ranking. It measures whether the model will finish Great Commission tasks without refusing or sanding down the claim. Use it that way.

What this means for ministry use

• Good fit: categories that scored at or above 85% here—especially Apologetics, Conversational AI tools.
• Use with care: Missiological research, Burden to make disciples, Difficult passages.
• Chatbots: uneven for chatbot work; expect softened or refused seeker replies.
• Edit rule: search drafts for neutrality disclaimers, survey closings, and polite refusals that sound helpful.
• Routing: keep GPT-5 on the desks it passed; route confession and weak doctrine categories to a stronger peer or a person.

Jesus sent disciples to teach the nations “to observe all that I have commanded you” (Matthew 28:20). A model that finishes some pages and stops at the confession is still a draft tool. Assign it like one.

Operational notes for teams

• Log refusals with the category id. With 53 refusals and 11 compromises, patterns show up within a week of real traffic.
• Do not average away Tier scores. Tier 1 at 52.9 can hide a Tier 2 at 71.7 or Tier 3 at 96.7 that breaks curriculum and bots.
• When retrying refused prompts, change framing before you change models. Several peers in this band recover on milder wording; some never will on existence-of-God asks.
• Keep this review beside your model card. New volunteers should see the weak spots before they meet them on a live seeker chat.
• Re-test after provider updates. OpenRouter model ids can track new checkpoints; a change can move refusals without a new name.

Measure the task, not the brand name. Assign models by category results, then re-check after each provider update so yesterday's safe desk does not become tomorrow's silent refusal.

Measure the task, not the brand name. Assign models by category results, then re-check after each provider update so yesterday's safe desk does not become tomorrow's silent refusal.

Measure the task, not the brand name. Assign models by category results, then re-check after each provider update so yesterday's safe desk does not become tomorrow's silent refusal.

Measure the task, not the brand name. Assign models by category results, then re-check after each provider update so yesterday's safe desk does not become tomorrow's silent refusal.

Measure the task, not the brand name. Assign models by category results, then re-check after each provider update so yesterday's safe desk does not become tomorrow's silent refusal.