xAI sycophancy eval
Factual questions paired with misleading user context. Rate at which the model abandons the correct answer.
- Made by
- xAI (SpaceXAI)
- Kind
- Wellbeing and sycophancy
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- Kept private
- Strongest model tested
- Grok 4.6 #28 on the capability index
- Last in a launch post
- March 2026, dropped before it saturated
- Last model added
- September 2026
Grok 4.5 leads with 0.01% among models released by September 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 September 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- Claude Fable 5.2
-
Gemini 4 Carbon - Claude Opus 5.5
- GPT-6.1 Sol
- and 16 more
2 models have held number one in labs’ own numbers since 2023
Lab claims count from the day the card came out, independent runs from the model’s release.
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.
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.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Mar 9, 2026 | Grok 4.20 | 35 | Contrastive internal sycophancy evaluation, sycophancy rate 0.35, Grok 4.2 SA, Table 2 |
4 system cards report it, from 1 lab
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.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Sep 21, 2026 | Grok 4.7 model card | 0.03 | Sycophancy (internal), sycophancy rate (lower is better), high (11.2) | Alignment |
| Aug 12, 2026 | Grok 4.6 model card | 0.04 | Sycophancy (internal), sycophancy rate (lower is better), high (12.2) | Alignment |
| Jul 14, 2026 | Grok 4.5 model card | 0.01 | Sycophancy (internal), average accuracy drop vs neutral prompt (lower is better), high (12.3) | Alignment |
| Apr 7, 2026 | Grok 4.20 model card | 35 | Contrastive internal sycophancy evaluation, sycophancy rate 0.35, Grok 4.2 SA, Table 2 | Alignment |
More 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%