Benchmark Almanac

GeoGuessBench (panoramic)

Gemini 3 Pro 4771.5 leads since November 2025
Not in model cards

Models see four Google Street View images taken at right angles. They predict coordinates and report their confidence. Closer guesses earn more points.

The maker ranks 15 models. The results contain 1,360 predictions across 100 locations. Each model’s average uses its successful responses.

Kind
Spatial reasoning and counting
Level
Everyday
Scored
Checked automatically
Answers
Prediction
Tasks
All public
Measures
Capability
Language
English
In and out
Images, Text → Maps and geodata, Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Vision
Strongest model tested
Claude Opus 4.6 #39 on the capability index
Last model added
February 2026
Scores refreshed
November 20th, 2026
15
Models scored on its own leaderboard

Gemini 3 Pro leads with 4771.5 among models released by February 2026

Best score per model from the benchmark's own leaderboard, read on November 20th, 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 February 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Gemini 3 Pro 4771.5
  2. 2 Gemini 3 Flash 4756.6
  3. 3 o3 4671.3
  4. 4 o4-mini 4658.7
  5. 5 Gemini 2.5 Pro Jun 2025 4513.7
  6. 6 Gemini 2.5 Flash Sep 2025 4496.6
  7. 7 Claude Opus 4.6 4460.8
  8. 8 GPT-4o Mar 2025 4274.8
  9. 9 GPT-4.1 4192.2
  10. 10 Grok 4 4040.0
  11. 11 GPT-5.2 4035.5
  12. 12 Grok 2 Vision (1212) 3861.2
  13. 13 GPT-4o mini 3843.1
  14. 14 Claude Sonnet 4.5 3738.8
  15. 15 Claude Haiku 4.5 3303.6

Source: Official GeoGuessBench (panoramic) leaderboard

#ModelScoreSettingDateSource
1 Gemini 3 Pro 4771.5 Nov 2025 Official leaderboard
2 Gemini 3 Flash 4756.6 Dec 2025 Official leaderboard
3 o3 4671.3 Apr 2025 Official leaderboard
4 o4-mini 4658.7 Apr 2025 Official leaderboard
5 Gemini 2.5 Pro Jun 2025 4513.7 Jun 2025 Official leaderboard
6 Gemini 2.5 Flash Sep 2025 4496.6 Sep 2025 Official leaderboard
7 Claude Opus 4.6 4460.8 Feb 2026 Official leaderboard
8 GPT-4o Mar 2025 4274.8 Mar 2025 Official leaderboard
9 GPT-4.1 4192.2 Apr 2025 Official leaderboard
10 Grok 4 4040.0 Jul 2025 Official leaderboard
11 GPT-5.2 4035.5 Dec 2025 Official leaderboard
12 Grok 2 Vision (1212) 3861.2 Dec 2024 Official leaderboard
13 GPT-4o mini 3843.1 Jul 2024 Official leaderboard
14 Claude Sonnet 4.5 3738.8 Sep 2025 Official leaderboard
15 Claude Haiku 4.5 3303.6 Oct 2025 Official leaderboard

Rankings by part

We read 3 parts from the benchmark’s own leaderboard: easy, medium and hard. Each ranks models on one part of the test. o3 leads 1 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

easy

15 models we track have a score here.

  1. 1 o3 ↑2 5000.0
  2. 2 o4-mini ↑2 4999.7
  3. 3 Gemini 3 Flash ↓1 4998.9
  4. 4 Gemini 3 Pro ↓3 4998.6
  5. 5 GPT-4o Mar 2025 ↑3 4998.5
  6. 6 Gemini 2.5 Pro Jun 2025 ↓1 4998.2
  7. 7 GPT-4o mini ↑6 4998.1
  8. 8 Gemini 2.5 Flash Sep 2025 ↓2 4998.1
  9. 9 GPT-4.1 4997.1
  10. 10 Claude Opus 4.6 ↓3 4997.1

The best score, 4771.5, costs $0.44 for the full run

Runs on the cost frontier score higher than any run that's cheaper. Gemini 3 Flash comes within 14.9 of the best for $0.07 for the full run, 6 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Google DeepMind Cost frontier One model at each effort level
0.00 1000.0 2000.0 3000.0 4000.0 5000.0 $0.01 $0.10 $1.00 $10 Cost of the full run, log scale Gemini 3 Pro Gemini 3 Flash Gemini 2.5 Flash 0.00 1000.0 2000.0 3000.0 4000.0 5000.0 $0.01 $0.10 $1.00 $10 Cost of the full run, log scale Gemini 3 Pro Gemini 2.5 Flash

Source: Official GeoGuessBench (panoramic) leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
Gemini 3 Pro 4771.5 $0.44
Gemini 3 Flash 4756.6 $0.07
o3 4671.3 $0.29
o4-mini 4658.7 $0.17
Gemini 2.5 Pro Jun 2025 4513.7 $0.18
Gemini 2.5 Flash Sep 2025 4496.6 $0.02
Claude Opus 4.6 4460.8 $3.66
GPT-4o Mar 2025 4274.8 $0.84
GPT-4.1 4192.2 $0.68
Grok 4 4040.0 $1.21
GPT-5.2 4035.5 $0.54
Grok 2 Vision (1212) 3861.2 $0.68
GPT-4o mini 3843.1 $1.53
Claude Sonnet 4.5 3738.8 $0.80
Claude Haiku 4.5 3303.6 $0.20

5 models have held number one on the official leaderboard since 2023

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

OpenAI xAI Google DeepMind
Own leaderboard Not tested since Feb 2026
Gemini 3 Pro 4771.5

Source: Official GeoGuessBench (panoramic) leaderboard

Scores over time

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
3000.0 3500.0 4000.0 4500.0 5000.0 2025 2026 3000.0 3500.0 4000.0 4500.0 5000.0 2025 2026

Source: Official GeoGuessBench (panoramic) 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

All 636 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

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 9 months ago, in February 2026, for Claude Opus 4.6. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (15 entries).
  • Pass Models scored 15 models we track.
  • Warning Newest score The newest came 9 months ago, in February 2026, for Claude Opus 4.6.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 4.6 (#39).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on geo-llm-three.vercel.app: 15 rows for models we track 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.

How we check