StrataBench Land v1 Extract
Qwen3-32B 88.1% leads since November 2026 Models extract ten shared lease fields from 628 synthetic instruments. Collide divides correct field decisions by correct, omitted, and wrong decisions, excluding unscored fields.
One benchmark of StrataBench Land v1. The suite publishes no combined score. Collide reports natural and schema prompts on the same ten fields. Collide keeps the schema protocol’s additional four fields outside the headline score.
- Made by
- Collide AI
- Released
- November 2026
- Kind
- Law
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Open-ended
- Tasks
- Public, plus a private test set
- Strongest model tested
- GPT-5.6 Luna #35 on the capability index
- Last model added
- July 2026
- Scores refreshed
- November 8th, 2026
Qwen3-32B leads with 88.1% among models released by July 2026
Best score per model from the benchmark's own leaderboard, read on November 8th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
The newest model with a score here came out in July 2026, and nobody has tested 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 - GPT-6 Astra
- Claude Opus 5.5
- and 16 more
- 1 Qwen3-32B schema 88.1%
- 2 GLM-5 schema 88%
- 3 Kimi K2.5 schema 87.9%
- 4 GPT-5.6 Luna schema 85.9%
- 5
Llama 3.3 70B schema 85.5%
Source: Official StrataBench Land v1 Extract leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Qwen3-32B | 88.1% | schema | Apr 2025 | Official leaderboard |
| 2 | GLM-5 | 88% | schema | Feb 2026 | Official leaderboard |
| 3 | Kimi K2.5 | 87.9% | schema | Jan 2026 | Official leaderboard |
| 4 | GPT-5.6 Luna | 85.9% | schema | Jul 2026 | Official leaderboard |
| 5 | Llama 3.3 70B | 85.5% | schema | Dec 2024 | 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.
Qwen3-32B Llama 3.3 70B Source: Official StrataBench Land v1 Extract leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Qwen3-32B | schema | 88.1% | |||
| GLM-5 | schema | 88% | |||
| Kimi K2.5 | schema | 87.9% | |||
| GPT-5.6 Luna | schema | 85.9% | |||
| Llama 3.3 70B | schema | 85.5% | |||
| GLM-5 | natural | 46% | |||
| Llama 3.3 70B | natural | 41.5% | |||
| Kimi K2.5 | natural | 33.4% | |||
| GPT-5.6 Luna | natural | 19% | |||
| Qwen3-32B | natural | 3.70% |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official StrataBench Land v1 Extract leaderboard
Llama 3.3 70B on the official leaderboard passed 85% in December 2024
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official StrataBench Land v1 Extract leaderboard
No lab model card reports it yet
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.
More in Law, finance and medicine
- MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy. Qwen 3.8 Max 80.4%
- HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
- HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
- FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy.
Ling-3.1-flash 81.7% - KBL In use Models answer Korean legal knowledge and reasoning tasks and Korean bar exam questions. It has 7 knowledge tasks, 4 reasoning tasks and 2,510 bar exam items.
Solar Pro 4 77.5% - KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. Solar Pro 4 93.2%