BigEarthNet.txt
Models answer land cover questions, caption satellite images, and locate regions described in text. The authors checked annotations for 1,082 pairs of radar and optical images.
Axis Spatial lists 22 configurations on a fixed sample of 240 questions. Its score excludes captioning and localization. The authors publish separate scores for all four tasks.
- Released
- March 2026
- Kind
- General visual questions
- Level
- University
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Capability
- Subject
- Earth and environmental sciences
- Language
- English
- In and out
- Text, Images, Maps and geodata → Text
- Interaction
- Single turn
- Items from
- Another benchmark, Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- Based on
- A version of another benchmark
- In views
- Vision
- Strongest model tested
- GPT-5.2 #51 on the capability index
- Last model added
- December 2025
- Scores refreshed
- November 20th, 2026
InternVL * (ours) tops the leaderboard with 34% GLM-4.6V-Flash leads with 0.79% among models released by December 2025
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is GLM-4.6V-Flash, at 0.79%. This view shows only models we track, each at its best.
The newest model with a score here came out in December 2025, 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 InternVL * (ours) 34%
- 2 EarthMind rgb 1.66%
- 3 LLaVa 0.96%
- 4 GLM-4.6V-Flash 0.79%
- 5 GeoChat 0.75%
- 6 LHRS-Bot 0.72%
- 7 Qwen 0.70%
- 8 EarthDial rgb 0.48%
- 9
InternVL3-1B 0.34%
- 10 GPT-5.2 0.30%
- 11 SkyEyeGPT 0%
Source: Official BigEarthNet.txt leaderboard
- 1 GLM-4.6V-Flash 0.79%
- 2
InternVL3-1B 0.34%
- 3 GPT-5.2 0.30%
Source: Official BigEarthNet.txt leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | InternVL * (ours) | 34% | Official leaderboard | ||
| 2 | EarthMind 4B | 1.66% | rgb | Official leaderboard | |
| 3 | LLaVa 7B | 0.96% | Official leaderboard | ||
| 4 | GLM-4.6V-Flash | 0.79% | Dec 2025 | Official leaderboard | |
| 5 | GeoChat 7B | 0.75% | Official leaderboard | ||
| 6 | LHRS-Bot 7B | 0.72% | Official leaderboard | ||
| 7 | Qwen 8B | 0.70% | Official leaderboard | ||
| 8 | EarthDial 4B | 0.48% | rgb | Official leaderboard | |
| 9 | | 0.34% | Apr 2025 | Official leaderboard | |
| 10 | GPT-5.2 | 0.30% | Dec 2025 | Official leaderboard | |
| 11 | SkyEyeGPT 7B | 0% | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GLM-4.6V-Flash | 0.79% | Dec 2025 | Official leaderboard | |
| 2 | | 0.34% | Apr 2025 | Official leaderboard | |
| 3 | GPT-5.2 | 0.30% | Dec 2025 | Official leaderboard |
Rankings by part
We read 4 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Captioning, Binary VQA and MCQ. GPT-5.2 leads 3 of them, and GLM-4.6V-Flash leads the other one. Arrows show how far a model moved from its overall rank among the models we track.
Captioning
3 models we track have a score here.
- 1 GLM-4.6V-Flash 0.79%
- 2
InternVL3-1B 0.34%
- 3 GPT-5.2 0.30%
Binary VQA
3 models we track have a score here.
- 1 GPT-5.2 ↑2 60.4%
- 2 GLM-4.6V-Flash ↓1 56.6%
- 3
InternVL3-1B ↓1 54.1%
MCQ
3 models we track have a score here.
- 1 GPT-5.2 ↑2 34.9%
- 2 GLM-4.6V-Flash ↓1 34.5%
- 3
InternVL3-1B ↓1 26.8%
Ref. Exp. Detection
3 models we track have a score here.
- 1 GPT-5.2 ↑2 31.7%
- 2 GLM-4.6V-Flash ↓1 27.2%
- 3
InternVL3-1B ↓1 5.76%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official BigEarthNet.txt leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official BigEarthNet.txt 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 Vision
- MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy.
Qwen 3.6 Plus 86%
- MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
- CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
- MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy.
Seed2.1 Pro 90.7%
- MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy.
Qwen 3.8 Max 97.7%
- MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy.
Qwen 3.8 Max 85.9%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 11 months ago, in December 2025, for GPT-5.2. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the official leaderboard (11 entries).
- Pass Models scored 3 models we track, plus 8 other entries on its leaderboard.
- Warning Newest score The newest came 11 months ago, in December 2025, for GPT-5.2.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.2 (#51).
- Pass Headroom The best is 0.79%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on txt.bigearth.net: 3 rows for models we track and 8 other entries at the last read.
- Pass Last read November 20th, 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.