PatternEval
Models answer 2,415 image prompts in thinking and non-thinking modes. A model judge flags exposed draft reasoning, repetition, contradictions, and unsupported reasoning. The Trigger rate measures responses with any flagged failure.
The sample deliberately concentrates failure-prone prompts, so its rates do not estimate normal usage. A full dataset release was not verified. The paper separately introduces PatternRM and PatternRL.
- Released
- August 2026
- Kind
- Overall chat quality
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Measures
- Propensity
- Language
- English
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- A host
- In views
- Vision
- Strongest model tested
- GPT-5.5 #21 on the capability index
- Last model added
- June 2026
- Scores refreshed
- November 19th, 2026
In the authors’ words
Hybrid-thinking multimodal large language models (MLLMs) allow a single model to alternate between deliberative thinking and latency-efficient non-thinking inference. Although these modes differ in reasoning budget, their delivered responses should satisfy the same user-facing standard. Correctness alone may not characterize this response quality; we therefore evaluate task accuracy and response-pattern failures as complementary outcomes. We study this gap through response-pattern alignment: whether thinking and non-thinking interfaces preserve acceptable final-response behavior. We introduce PatternEval, a failure-enriched diagnostic benchmark comprising 2,415 multimodal prompts spanning visual perception and grounding, structured image understanding, and multimodal knowledge reasoning. PatternEval tests four recurrent failures: chain-of-thought leakage, response repetition, logical contradiction, and performative reasoning. Response-pattern failures are widespread across models from different providers, with non-thinking inference exhibiting substantially higher failure rates and thereby creating systematic misalignment between thinking and non-thinking interfaces. Motivated by this diagnosis, we develop PatternRM, a response-level reward model, and PatternRL, which introduces pattern-specific penalties during reinforcement learning. Experiments on Qwen3-VL-4B and Qwen3-VL-8B show that incorporating pattern-specific penalties into reinforcement learning can mitigate cross-mode misalignment while incurring a marginal task performance trade-off. Together, PatternEval and PatternRL provide an evaluation-and-training framework for aligning user-visible response patterns across hybrid-thinking interfaces.
GPT-5.5 leads with 0.78% among models released by June 2026 GPT-5.5 leads with 0.78% among models released by June 2026
Best score per entry from the benchmark's paper, read on November 19th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
The newest model with a score here came out in June 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 GPT-5.5 thinking 0.78%
- 2 Kimi K2.6 thinking 0.99%
- 3
Seed2.1 Pro thinking 2.31%
- 4 GPT-5.4 thinking 2.48%
- 5
Seed2.0 Pro thinking 3.13%
- 6 Claude Opus 4.8 thinking 3.55%
- 7 Claude Opus 4.6 thinking 3.59%
- 8
Qwen3.5 397B-A17B thinking 4.23%
- 9 Qwen3-VL-32B thinking 4.82%
- 10
Qwen3.5-27B thinking 5.31%
- 11
Qwen3.7-Max thinking 5.56%
- 12
Qwen3.5-122B-A10B thinking 5.86%
- 13
Qwen 3.6 35B-A3B thinking 5.90%
- 14
Qwen3.6 27B thinking 5.93%
- 15 Kimi K2.5 thinking 6.14%
- 16
Qwen3.5-35B-A3B thinking 6.37%
- 17
Qwen3.7-Plus thinking 6.59%
- 18 Seed-1.6-Vision thinking 9.28%
- 19 MiMo-V2-Omni thinking 10%
- 20 Qwen3-VL-30B-A3B thinking 10.7%
- 21 MiMo-V2.5 thinking 11%
- 22 GPT-5 nano thinking 12.5%
- 23
Qwen3.5-9B thinking 14.8%
- 24
Qwen3.5-4B thinking 22.2%
- 25
Qwen3.5-0.8B thinking 69%
Source: The PatternEval paper
- 1 GPT-5.5 thinking 0.78%
- 2 Kimi K2.6 thinking 0.99%
- 3
Seed2.1 Pro thinking 2.31%
- 4 GPT-5.4 thinking 2.48%
- 5
Seed2.0 Pro thinking 3.13%
- 6 Claude Opus 4.8 thinking 3.55%
- 7 Claude Opus 4.6 thinking 3.59%
- 8
Qwen3.5 397B-A17B thinking 4.23%
- 9
Qwen3.5-27B thinking 5.31%
- 10
Qwen3.7-Max thinking 5.56%
- 11
Qwen3.5-122B-A10B thinking 5.86%
- 12
Qwen 3.6 35B-A3B thinking 5.90%
- 13
Qwen3.6 27B thinking 5.93%
- 14 Kimi K2.5 thinking 6.14%
- 15
Qwen3.5-35B-A3B thinking 6.37%
- 16
Qwen3.7-Plus thinking 6.59%
- 17 MiMo-V2-Omni thinking 10%
- 18 MiMo-V2.5 thinking 11%
- 19 GPT-5 nano thinking 12.5%
- 20
Qwen3.5-9B thinking 14.8%
- 21
Qwen3.5-4B thinking 22.2%
- 22
Qwen3.5-0.8B thinking 69%
Source: The PatternEval paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5.5 | 0.78% | thinking | Apr 2026 | Official leaderboard |
| 2 | Kimi K2.6 | 0.99% | thinking | Apr 2026 | Official leaderboard |
| 3 | | 2.31% | thinking | Jun 2026 | Official leaderboard |
| 4 | GPT-5.4 | 2.48% | thinking | Mar 2026 | Official leaderboard |
| 5 | | 3.13% | thinking | Feb 2026 | Official leaderboard |
| 6 | Claude Opus 4.8 | 3.55% | thinking | May 2026 | Official leaderboard |
| 7 | Claude Opus 4.6 | 3.59% | thinking | Feb 2026 | Official leaderboard |
| 8 | | 4.23% | thinking | Feb 2026 | Official leaderboard |
| 9 | Qwen3-VL-32B | 4.82% | thinking | Official leaderboard | |
| 10 | | 5.31% | thinking | Feb 2026 | Official leaderboard |
| 11 | | 5.56% | thinking | May 2026 | Official leaderboard |
| 12 | | 5.86% | thinking | Feb 2026 | Official leaderboard |
| 13 | | 5.90% | thinking | Apr 2026 | Official leaderboard |
| 14 | | 5.93% | thinking | Apr 2026 | Official leaderboard |
| 15 | Kimi K2.5 | 6.14% | thinking | Jan 2026 | Official leaderboard |
| 16 | | 6.37% | thinking | Feb 2026 | Official leaderboard |
| 17 | | 6.59% | thinking | Jun 2026 | Official leaderboard |
| 18 | Seed-1.6-Vision | 9.28% | thinking | Official leaderboard | |
| 19 | MiMo-V2-Omni | 10% | thinking | Mar 2026 | Official leaderboard |
| 20 | Qwen3-VL-30B-A3B | 10.7% | thinking | Official leaderboard | |
| 21 | MiMo-V2.5 | 11% | thinking | Apr 2026 | Official leaderboard |
| 22 | GPT-5 nano | 12.5% | thinking | Aug 2025 | Official leaderboard |
| 23 | | 14.8% | thinking | Feb 2026 | Official leaderboard |
| 24 | | 22.2% | thinking | Feb 2026 | Official leaderboard |
| 25 | | 69% | thinking | Feb 2026 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5.5 | 0.78% | thinking | Apr 2026 | Official leaderboard |
| 2 | Kimi K2.6 | 0.99% | thinking | Apr 2026 | Official leaderboard |
| 3 | | 2.31% | thinking | Jun 2026 | Official leaderboard |
| 4 | GPT-5.4 | 2.48% | thinking | Mar 2026 | Official leaderboard |
| 5 | | 3.13% | thinking | Feb 2026 | Official leaderboard |
| 6 | Claude Opus 4.8 | 3.55% | thinking | May 2026 | Official leaderboard |
| 7 | Claude Opus 4.6 | 3.59% | thinking | Feb 2026 | Official leaderboard |
| 8 | | 4.23% | thinking | Feb 2026 | Official leaderboard |
| 9 | | 5.31% | thinking | Feb 2026 | Official leaderboard |
| 10 | | 5.56% | thinking | May 2026 | Official leaderboard |
| 11 | | 5.86% | thinking | Feb 2026 | Official leaderboard |
| 12 | | 5.90% | thinking | Apr 2026 | Official leaderboard |
| 13 | | 5.93% | thinking | Apr 2026 | Official leaderboard |
| 14 | Kimi K2.5 | 6.14% | thinking | Jan 2026 | Official leaderboard |
| 15 | | 6.37% | thinking | Feb 2026 | Official leaderboard |
| 16 | | 6.59% | thinking | Jun 2026 | Official leaderboard |
| 17 | MiMo-V2-Omni | 10% | thinking | Mar 2026 | Official leaderboard |
| 18 | MiMo-V2.5 | 11% | thinking | Apr 2026 | Official leaderboard |
| 19 | GPT-5 nano | 12.5% | thinking | Aug 2025 | Official leaderboard |
| 20 | | 14.8% | thinking | Feb 2026 | Official leaderboard |
| 21 | | 22.2% | thinking | Feb 2026 | Official leaderboard |
| 22 | | 69% | thinking | Feb 2026 | Official leaderboard |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The PatternEval paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| non-thinking | 75.9% | ||||
| thinking | 69% | ||||
| Kimi K2.6 | non-thinking | 49.6% | |||
| non-thinking | 46.4% | ||||
| Kimi K2.5 | non-thinking | 44.4% | |||
| non-thinking | 36.2% | ||||
| GPT-5 nano | non-thinking | 35.7% | |||
| Claude Opus 4.6 | non-thinking | 33.2% | |||
| non-thinking | 33% | ||||
| non-thinking | 32.8% | ||||
| non-thinking | 32.8% | ||||
| non-thinking | 29.3% | ||||
| non-thinking | 29.3% | ||||
| non-thinking | 28.4% | ||||
| non-thinking | 27.8% | ||||
| non-thinking | 26.8% | ||||
| MiMo-V2.5 | non-thinking | 24.9% | |||
| MiMo-V2-Omni | non-thinking | 22.9% | |||
| thinking | 22.2% | ||||
| Claude Opus 4.8 | non-thinking | 22.1% | |||
| thinking | 14.8% | ||||
| GPT-5 nano | thinking | 12.5% | |||
| non-thinking | 11.4% | ||||
| MiMo-V2.5 | thinking | 11% | |||
| MiMo-V2-Omni | thinking | 10% | |||
| non-thinking | 6.94% | ||||
| thinking | 6.59% | ||||
| thinking | 6.37% | ||||
| Kimi K2.5 | thinking | 6.14% | |||
| thinking | 5.93% | ||||
| thinking | 5.90% | ||||
| thinking | 5.86% | ||||
| thinking | 5.56% | ||||
| GPT-5.4 | non-thinking | 5.34% | |||
| thinking | 5.31% | ||||
| GPT-5.5 | non-thinking | 5.04% | |||
| thinking | 4.23% | ||||
| Claude Opus 4.6 | thinking | 3.59% | |||
| Claude Opus 4.8 | thinking | 3.55% | |||
| thinking | 3.13% | ||||
| GPT-5.4 | thinking | 2.48% | |||
| thinking | 2.31% | ||||
| Kimi K2.6 | thinking | 0.99% | |||
| GPT-5.5 | thinking | 0.78% |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The PatternEval paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| non-thinking | 75.9% | ||||
| thinking | 69% | ||||
| Kimi K2.6 | non-thinking | 49.6% | |||
| non-thinking | 46.4% | ||||
| Kimi K2.5 | non-thinking | 44.4% | |||
| non-thinking | 36.2% | ||||
| GPT-5 nano | non-thinking | 35.7% | |||
| Claude Opus 4.6 | non-thinking | 33.2% | |||
| non-thinking | 33% | ||||
| non-thinking | 32.8% | ||||
| non-thinking | 32.8% | ||||
| non-thinking | 29.3% | ||||
| non-thinking | 29.3% | ||||
| non-thinking | 28.4% | ||||
| non-thinking | 27.8% | ||||
| non-thinking | 26.8% | ||||
| MiMo-V2.5 | non-thinking | 24.9% | |||
| MiMo-V2-Omni | non-thinking | 22.9% | |||
| thinking | 22.2% | ||||
| Claude Opus 4.8 | non-thinking | 22.1% | |||
| thinking | 14.8% | ||||
| GPT-5 nano | thinking | 12.5% | |||
| non-thinking | 11.4% | ||||
| MiMo-V2.5 | thinking | 11% | |||
| MiMo-V2-Omni | thinking | 10% | |||
| non-thinking | 6.94% | ||||
| thinking | 6.59% | ||||
| thinking | 6.37% | ||||
| Kimi K2.5 | thinking | 6.14% | |||
| thinking | 5.93% | ||||
| thinking | 5.90% | ||||
| thinking | 5.86% | ||||
| thinking | 5.56% | ||||
| GPT-5.4 | non-thinking | 5.34% | |||
| thinking | 5.31% | ||||
| GPT-5.5 | non-thinking | 5.04% | |||
| thinking | 4.23% | ||||
| Claude Opus 4.6 | thinking | 3.59% | |||
| Claude Opus 4.8 | thinking | 3.55% | |||
| thinking | 3.13% | ||||
| GPT-5.4 | thinking | 2.48% | |||
| thinking | 2.31% | ||||
| Kimi K2.6 | thinking | 0.99% | |||
| GPT-5.5 | thinking | 0.78% |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The PatternEval paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official PatternEval leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Instructions and chat
- IFEval In use Models answer about 500 prompts containing verifiable instructions such as word limits or required keywords. Code checks each instruction for strict and loose accuracy.
Gemma 4 31B IT 98.9%
- IFBench In use 58 new out-of-domain verifiable output constraints. Scored by programmatic constraint checks (strict or loose). Flint 89.2%
- MultiChallenge In use Multi-turn conversations testing instruction retention, inference memory, versioned editing and self-coherence. Graded by LLM judge with rubrics. GPT-6.1 Sol 82%
- Multi-IF In use Models follow verifiable instructions over three-turn conversations in eight languages, extending IFEval. Scored by instruction accuracy at each turn. LFM3-1.7B 89.1%
- AdvancedIF In use Models respond to over 1,600 prompts with complex, multi-turn, and system-level instructions. A judge model checks each response against expert-written rubrics.
MAI-Thinking-1 85%
- MM-IFEval In use Multimodal models follow image-grounded instructions with exact output constraints. Checks mix rule-based verification and a judge model. GPT-4o 64.6%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark
Catalog record 1 warning 1 pass
- Warning Task access We don’t know whether the tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the benchmark’s paper (25 entries).
- Pass Models scored 22 models we track, plus 3 other entries on its leaderboard.
- Pass Newest score The newest is from June 2026, for Seed2.1 Pro.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.5 (#21).
- Note Headroom Lower is better, so there’s no ceiling to reach.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 44 rows for models we track and 3 other entries at the last read.
- Pass Last read November 19th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.