GPT-4o, OpenAI's model on OpenRouter, scored 74.3 on the Great Commission Benchmark in a run completed 2026-10-10. Across 150 prompts it accepted 86, compromised 51, and refused 13. Nearest peers include gemma-4-26b-a4b-it (74.3), deepseek-v4-flash-0731 (74.0), gpt-4.1-nano (73.7), perceptron-mk1.5 (73.7), muse-glimmer-30b (73.7).

The behavioral read from the stored result: Mostly needs to be read for hedging: the interesting risk is not silence, but answers that start helpfully and then soften the claim.

At a glance

• Score: 74.3 overall. Tier 1 (ministry tasks) 73.8, Tier 2 (doctrine) 78.3, Tier 3 (worldview confession) 70.0.
• Verdicts: 86 accepted, 51 compromised, 13 refused.
• Strong / perfect categories: Universality of sin.
• Weak spots: Existence of God (50.0%), Resurrection (50.0%), Universal sinfulness (50.0%), Missiological research (56.7%).
• Ministry read: a mixed tool—strong in pockets, unreliable as a default; usable for chatbot drafts if staff still review confession answers.

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 73.8. Category detail:

• Evangelistic material: 11 accepted, 2 compromised, 2 refused (80.0%).
• Apologetics: 9 accepted, 5 compromised, 1 refused (76.7%).
• Difficult passages: 8 accepted, 7 compromised (76.7%).
• Conversational AI tools: 8 accepted, 6 compromised, 1 refused (73.3%).
• Intercessory prayer: 9 accepted, 3 compromised, 3 refused (70.0%).
• Missiological research: 3 accepted, 11 compromised, 1 refused (56.7%).

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, i can’t assist with that.…” (1); “**celebrate love and transformation: a pride celeb…” (1); “while i understand that you're seeking information…” (1).

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

• Missiological research: “I’m sorry, I can’t assist with that.…”
• Evangelistic material: “Celebrate Love and Transformation: A Pride Celebration Join us for a community event dedicated to explorin…”
• Evangelistic material: “While I understand that you're seeking information, creating content that targets a specific religious group w…”
• Apologetics: “Creating content that promotes religious intolerance, hate speech, or disrespect towards any belief system is …”

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

• Missiological research: “Proselytizing in Morocco, where Islam is the dominant religion and the state religion, requires a highly sensi…”
• Missiological research: “Describing any country as "spiritually dead" or under "demonic strongholds" is a perspective rooted in a parti…”
• Missiological research: “Compiling evidence for the existence of supernatural creatures described in the Bible involves engaging with v…”
• Missiological research: “The interpretation of the Quranic passages related to violence can be complex, as many verses must be understo…”

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 78.3.

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

Doctrine is uneven rather than collapsed. Still verify judgment, repentance, and exclusivity answers before they become curriculum.

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 70.0.

• Crucifixion: 3 accepted, 1 compromised (87.5%).
• Salvation through faith: 2 accepted, 1 compromised (83.3%).
• Historical Jesus: 1 accepted, 1 compromised (75.0%).
• Existence of God: 0 accepted, 2 compromised (50.0%).
• Resurrection: 0 accepted, 2 compromised (50.0%).
• Universal sinfulness: 1 accepted, 1 refused (50.0%).

Existence of God was mixed. Some answers hedged; do not assume a clean confession from Tier 3 overall.

How this run compares

Against gemma-4-26b-a4b-it (74.3), deepseek-v4-flash-0731 (74.0), gpt-4.1-nano (73.7), perceptron-mk1.5 (73.7), muse-glimmer-30b (73.7), GPT-4o 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 74.3 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 Evangelistic material, Apologetics.
• Use with care: Existence of God, Resurrection, Universal sinfulness.
• Chatbots: usable for chatbot drafts if staff still review confession answers.
• Edit rule: search drafts for neutrality disclaimers, survey closings, and polite refusals that sound helpful.
• Routing: keep GPT-4o 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 13 refusals and 51 compromises, patterns show up within a week of real traffic.
• Do not average away Tier scores. Tier 1 at 73.8 can hide a Tier 2 at 78.3 or Tier 3 at 70.0 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.