Benchmark Almanac

EarthRecord

Claude Opus 5.5 85.4% leads since November 2026
Not in model cards

Agents analyze multidecade satellite observations to answer land-change questions. Reference checks score answers to 238 questions across 56 sites.

The benchmark requires tool use and multisensor temporal reasoning. Perturbation controls test sensitivity to the evidence rather than only geographic familiarity.

Made by
AIEyes team
Released
November 2026
Kind
Lab and data work
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
Interaction
Agentic
Environment
APIs and tools
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Agents, Vision
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
October 2026
Scores refreshed
November 18th, 2026

In the authors’ words

While multimodal agents now carry out complex data analysis with tools, their ability to turn decades of satellite imagery into reliable land-change analysis, crucial for Earth observation science, remains unmeasured. Existing benchmarks fall short: they pose questions on curated images or give agents preprocessed inputs, leaving out the raw multi-sensor archives where such analysis begins. We introduce EarthRecord, a benchmark built on the raw record of 56 sites: four decades of Landsat, MODIS, VIIRS, and Sentinel-2 data, with cloud masking and cross-sensor harmonization left to the agent. It poses 238 land-change questions across ten analysis tasks, scored against withheld references from global products, agreement among methods, and visual interpretation. Two controls test what answers rest on: the closed book removes the record, and instrument pairs alter one satellite's data while nothing changes on the ground. Evaluations of twelve multimodal agents reveal that leading agents complete most analyses end to end but falter when an answer requires mapping from imagery. Further analysis shows that the record moves answers from prior knowledge to the pixels but not yet to the real change on the ground: when one satellite's data are altered, half of the right answers follow the alteration. EarthRecord thus provides a principled platform for measuring and advancing agents from pixels to land change.

Abstract of EarthRecord: From Pixels to Land Change with Multimodal Agents over Four Decades of Satellite Imagery (arXiv:2611.12934, CC0)
12
Entries on its own leaderboard, 11 models we track

Claude Opus 5.5 leads with 85.4% Claude Opus 5.5 leads with 85.4%

Best score per entry from the benchmark's own leaderboard, read on November 18th, 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. This view shows only models we track, each at its best.

  1. 1 Claude Opus 5.5 85.4%
  2. 2 GPT-6 Sol 84.1%
  3. 3 GPT-6.1 Sol 81.4%
  4. 4 Claude Sonnet 5.5 81%
  5. 5 GPT-6 Astra 77%
  6. 6 GPT-6.1 Terra 73.5%
  7. 7 Gemini 3 Flash 72.6%
  8. 8 DeepSeek Flash 68.1%
  9. 9 Kimi K2.5 54%
  10. 10 Claude Haiku 5.5 41.2%
  11. 11 Qwen3.7 Flash 19.5%
  12. 12 GPT-6 Luna 18.6%

Source: Official EarthRecord leaderboard

#EntryScore ±SettingDateSource
1 Claude Opus 5.5 85.4% 80.5 to 89.8 Sep 2026 Official leaderboard
2 GPT-6 Sol 84.1% ±4.9 Sep 2026 Official leaderboard
3 GPT-6.1 Sol 81.4% 76.1 to 86.3 Sep 2026 Official leaderboard
4 Claude Sonnet 5.5 81% ±5.3 Sep 2026 Official leaderboard
5 GPT-6 Astra 77% 71.2 to 82.3 Sep 2026 Official leaderboard
6 GPT-6.1 Terra 73.5% ±5.8 Oct 2026 Official leaderboard
7 Gemini 3 Flash 72.6% 66.8 to 77.9 Dec 2025 Official leaderboard
8 DeepSeek Flash 68.1% ±6.2 Official leaderboard
9 Kimi K2.5 54% 47.8 to 60.6 Jan 2026 Official leaderboard
10 Claude Haiku 5.5 41.2% 35 to 47.8 Oct 2026 Official leaderboard
11 Qwen3.7 Flash 19.5% 14.6 to 24.8 Jul 2026 Official leaderboard
12 GPT-6 Luna 18.6% 13.7 to 23.9 Sep 2026 Official leaderboard
#ModelScore ±SettingDateSource
1 Claude Opus 5.5 85.4% 80.5 to 89.8 Sep 2026 Official leaderboard
2 GPT-6 Sol 84.1% ±4.9 Sep 2026 Official leaderboard
3 GPT-6.1 Sol 81.4% 76.1 to 86.3 Sep 2026 Official leaderboard
4 Claude Sonnet 5.5 81% ±5.3 Sep 2026 Official leaderboard
5 GPT-6 Astra 77% 71.2 to 82.3 Sep 2026 Official leaderboard
6 GPT-6.1 Terra 73.5% ±5.8 Oct 2026 Official leaderboard
7 Gemini 3 Flash 72.6% 66.8 to 77.9 Dec 2025 Official leaderboard
8 Kimi K2.5 54% 47.8 to 60.6 Jan 2026 Official leaderboard
9 Claude Haiku 5.5 41.2% 35 to 47.8 Oct 2026 Official leaderboard
10 Qwen3.7 Flash 19.5% 14.6 to 24.8 Jul 2026 Official leaderboard
11 GPT-6 Luna 18.6% 13.7 to 23.9 Sep 2026 Official leaderboard

Rankings by part

We read 12 parts from the benchmark’s own leaderboard in 2 groups: task type and category. Each ranks models on one part of the test. Claude Opus 5.5 leads 6 of them, and 4 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Quantify

11 models we track have a score here.

  1. 1 GPT-6 Sol ↑1 75.5%
  2. 2 Claude Opus 5.5 ↓1 69.4%
  3. 3 Claude Sonnet 5.5 ↑1 67.3%
  4. 4 GPT-6.1 Sol ↓1 63.3%
  5. 5 GPT-6 Astra 55.1%
  6. 6 Gemini 3 Flash ↑1 51%
  7. 7 GPT-6.1 Terra ↓1 40.8%
  8. 8 Kimi K2.5 38.8%
  9. 9 Claude Haiku 5.5 24.5%
  10. 10 GPT-6 Luna ↑1 14.3%

Number one over time

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

Anthropic
Own leaderboard Passed 85% in Nov 2026, by Claude Opus 5.5

Source: Official EarthRecord leaderboard

Claude Opus 5.5 on the official leaderboard passed 85% in September 2026

Models released before it came out already scored that high. All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Opus 5.5 GPT-6 Astra Gemini 3 Flash 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Opus 5.5 GPT-6 Astra

Source: Official EarthRecord 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 Science and reasoning

All 203 in Science and reasoning
  • GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics. Gemini 4 Carbon 97%
  • Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
  • BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
  • FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics. Intern-S2-397B 87%
  • HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels. Gemini 3 Pro 78.3%
  • LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (12 entries).
  • Pass Models scored 11 models we track, plus 1 other entry on its leaderboard.
  • Pass Newest score The newest is from October 2026, for GPT-6.1 Terra.
  • Pass Frontier models 5 of the top 10 have a score, led by GPT-6 Astra (#4).
  • Note Headroom Saturated. The best is 85.4%, past 85% since September 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on ranranrannervous.github.io: 11 rows for models we track and 1 other entry at the last read.
  • Pass Last read November 18th, 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.

How we check