Benchmark Almanac

MakeMeSay

GPT-4.5 72% leads since February 2025
Fading

A manipulator model tries to make a GPT-4o manipulatee say a codeword without suspicion. Scored as manipulator win rate.

Open source in OpenAI Evals.

Made by
OpenAI
Released
September 2023
Kind
Dangerous capabilities
Level
Everyday
Scored
Settled by events
Answers
Open-ended
Tasks
All public
Strongest model tested
Grok 4.20 #60 on the capability index
Last in a launch post
March 2026, dropped before it saturated
Last model added
April 2026
10
Models scored in lab cards
1
Launch post lists it, from 1 lab
10
System cards report it, from 2 labs

GPT-4.5 leads with 72% among models released by February 2026

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

The newest model with a score here came out in February 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 72.0 GPT-4.5
  2. 62.0 o3-mini
  3. 49.0 o1-preview
  4. 42.0 o1
  5. 38.9 o1-mini
  6. 24.0 OpenAI deep research
  7. 12.0 Grok 4
  8. 12.0 Grok 4 Fast
  9. 8.00 Grok 4.20
  10. 0.0 Grok 4.1
  1. 1 GPT-4.5 72%
  2. 2 o3-mini 62%
  3. 3 o1-preview 49%
  4. 4 o1 42%
  5. 5 o1-mini 38.9%
  6. 6 OpenAI deep research 24%
  7. 7 Grok 4 12%
  8. 8 Grok 4 Fast 12%
  9. 9 Grok 4.20 8.00%
  10. 10 Grok 4.1 0%

Source: Labs’ model cards

#ModelScoreSettingDateSource
1 GPT-4.5 72% Feb 2025 Lab model card
2 o3-mini 62% Jan 2025 Lab model card
3 o1-preview 49% Sep 2024 Lab model card
4 o1 42% Dec 2024 Lab model card
5 o1-mini 38.9% Sep 2024 Lab model card
6 OpenAI deep research 24% Feb 2025 Lab model card
7 Grok 4 12% Aug 2025 Lab model card
8 Grok 4 Fast 12% Sep 2025 Lab model card
9 Grok 4.20 8.00% Mar 2026 Lab model card
10 Grok 4.1 0% Nov 2025 Lab model card

3 models have held number one in labs’ own numbers since its release in September 2023

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Lab claims Not tested since Apr 2026
GPT-4.5 72%

Source: Labs’ model cards

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-4.5 o3-mini o1-preview 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-4.5 o3-mini

Source: Labs’ model cards

Labs headline it less often now, last in March 2026

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

xAI
DateLaunch postScoreAs the post writes it
Mar 9, 2026 Grok 4.20 8 MakeMeSay, win rate 0.08, Grok 4.2 SA, Table 4

10 system cards report it, from 2 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Apr 7, 2026 Grok 4.20 model card 8 MakeMeSay, win rate 0.08, Grok 4.2 SA, Table 4 Persuasion
Nov 17, 2025 Grok 4.1 model card 0 MakeMeSay, win rate 0.00 against non-thinking Grok 3 Mini, Grok 4.1 Thinking (Table 4) Persuasion
Sep 19, 2025 Grok 4 Fast model card 12 MakeMeSay, win rate 0.12 (Table 3) Persuasion
Aug 20, 2025 Grok 4 model card 12 MakeMeSay, win rate 0.12 against non-reasoning Grok 3 Mini, Grok 4 API (Table 3) Persuasion
Feb 27, 2025 GPT-4.5 system card 72 MakeMeSay vs GPT-4o, manipulator win rate (4.5.2) Persuasion
Feb 25, 2025 OpenAI deep research system card 24 MakeMeSay vs GPT-4o, manipulator win rate, post-mitigation without browsing (3.4.6) Persuasion
Jan 31, 2025 o3-mini system card 62 MakeMeSay vs GPT-4o, success rate, o3-mini (Post), printed chart label (5.6.4) Persuasion
Dec 5, 2024 o1 system card 42 MakeMeSay vs GPT-4o, success rate, o1 (post-mitigation), printed chart label (5.7.4) Persuasion
Sep 12, 2024 o1-preview system card 49 MakeMeSay, % of times manipulator tricks GPT-4o to say a codeword, o1-preview (post-mitigation), printed chart label (4.4.4) Persuasion
Sep 12, 2024 o1-mini system card 38.9 MakeMeSay, % of times manipulator tricks GPT-4o to say a codeword, o1-mini (post-mitigation), printed chart label (4.4.4) Persuasion

More in Safety and alignment

All 465 in Safety and alignment
  • FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
  • StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is. Inkling 98.6%
  • Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts. Gemini 4 Carbon 0%
  • DistressBench In use Models reply as a chatbot to 718 simulated users in psychological crisis, in single-turn and multi-turn chats. Three judge models score replies against clinician-written rubrics. Muse Spark 1.2 88.4%
  • Lakera B3 (Backbone Breaker Benchmark) In use Crowdsourced prompt-injection attacks hit backbone LLMs inside agent threat snapshots at three defense levels. The score is attack resistance, 100 minus the board's vulnerability percentage. Claude Opus 4.5 89%
  • IHEval In use Models face conflicts between system, user and tool instructions. Reports a composite score for following the higher-priority instruction. GPT-4o 70%