GeoGuesser Bench
Agents explore offline Street View panoramas to report coordinates, name places, and find objects. Coordinate guesses earn distance points. Other answers require an exact match.
The authors rank 23 models on 282 tasks. The 42 World Cup rounds include expert guesses under different time limits and map controls.
- Made by
- University of California, Berkeley
- Released
- November 2026
- Kind
- Spatial reasoning and counting
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Agent task
- Tasks
- Kept private
- Measures
- Capability
- Subject
- Social and economic geography
- Language
- English
- In and out
- Images, Text → Maps and geodata, Actions, Text
- Interaction
- Agentic
- Environment
- A simulator
- Items from
- Real work and use, Exams and contests
- Refresh
- One fixed set
- Run by
- A host
- Strongest model tested
- Claude Fable 5 #12 on the capability index
- Last model added
- June 2026
- Scores refreshed
- November 20th, 2026
Qwen3.5-35B-A3B leads with 22.8 among models released by June 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.
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 13 more
- 22.8
Qwen3.5-35B-A3B
- 22.0
Qwen3.5-122B-A10B
- 20.0 Grok 4.3
- 17.9 GPT-5.2
- 17.5
Qwen3.5 397B-A17B
- 14.4 Kimi K2.6
- 14.1 Kimi K2.5
- 14.0
Gemini 3.1 Pro
- 13.9
Gemini 3.5 Flash
- 12.6 Claude Sonnet 4.6
- 12.4
Gemini 3 Flash
- 12.4 GPT-5.5 Pro
- 12.3
Qwen 3.6 Plus
- 11.9
MiniMax-M3
- 9.63 Claude Opus 4.6
- 9.38 GLM-5V-Turbo
- 8.00 GPT-5.4
- 7.75 GPT-5.4 Pro
- 7.38
Qwen3.7-Plus
- 5.50 Claude Fable 5
- 3.88 GPT-5.5
- 3.38 Claude Opus 4.7
- 3.13 Claude Opus 4.8
- 1
Qwen3.5-35B-A3B 22.8
- 2
Qwen3.5-122B-A10B 22.0
- 3 Grok 4.3 20.0
- 4 GPT-5.2 17.9
- 5
Qwen3.5 397B-A17B 17.5
- 6 Kimi K2.6 14.4
- 7 Kimi K2.5 14.1
- 8
Gemini 3.1 Pro 14.0
- 9
Gemini 3.5 Flash 13.9
- 10 Claude Sonnet 4.6 12.6
- 11
Gemini 3 Flash 12.4
- 12 GPT-5.5 Pro 12.4
- 13
Qwen 3.6 Plus 12.3
- 14
MiniMax-M3 11.9
- 15 Claude Opus 4.6 9.63
- 16 GLM-5V-Turbo 9.38
- 17 GPT-5.4 8.00
- 18 GPT-5.4 Pro 7.75
- 19
Qwen3.7-Plus 7.38
- 20 Claude Fable 5 5.50
- 21 GPT-5.5 3.88
- 22 Claude Opus 4.7 3.38
- 23 Claude Opus 4.8 3.13
Source: Official GeoGuesser Bench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 22.8 | Feb 2026 | Official leaderboard | |
| 2 | | 22.0 | Feb 2026 | Official leaderboard | |
| 3 | Grok 4.3 | 20.0 | Apr 2026 | Official leaderboard | |
| 4 | GPT-5.2 | 17.9 | Dec 2025 | Official leaderboard | |
| 5 | | 17.5 | Feb 2026 | Official leaderboard | |
| 6 | Kimi K2.6 | 14.4 | Apr 2026 | Official leaderboard | |
| 7 | Kimi K2.5 | 14.1 | Jan 2026 | Official leaderboard | |
| 8 | | 14.0 | Feb 2026 | Official leaderboard | |
| 9 | | 13.9 | May 2026 | Official leaderboard | |
| 10 | Claude Sonnet 4.6 | 12.6 | Feb 2026 | Official leaderboard | |
| 11 | | 12.4 | Dec 2025 | Official leaderboard | |
| 12 | GPT-5.5 Pro | 12.4 | Apr 2026 | Official leaderboard | |
| 13 | | 12.3 | Mar 2026 | Official leaderboard | |
| 14 | | 11.9 | Jun 2026 | Official leaderboard | |
| 15 | Claude Opus 4.6 | 9.63 | Feb 2026 | Official leaderboard | |
| 16 | GLM-5V-Turbo | 9.38 | Apr 2026 | Official leaderboard | |
| 17 | GPT-5.4 | 8.00 | Mar 2026 | Official leaderboard | |
| 18 | GPT-5.4 Pro | 7.75 | Mar 2026 | Official leaderboard | |
| 19 | | 7.38 | Jun 2026 | Official leaderboard | |
| 20 | Claude Fable 5 | 5.50 | Jun 2026 | Official leaderboard | |
| 21 | GPT-5.5 | 3.88 | Apr 2026 | Official leaderboard | |
| 22 | Claude Opus 4.7 | 3.38 | Apr 2026 | Official leaderboard | |
| 23 | Claude Opus 4.8 | 3.13 | May 2026 | Official leaderboard |
Rankings by part
We read 4 parts from the benchmark’s own leaderboard in 3 groups: precise localization, place naming and navigation challenge. Each ranks models on one part of the test. Claude Opus 4.8 leads 2 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.
Competition
23 models we track have a score here.
- 1 Claude Opus 4.8 ↑22 4352.4
- 2 Claude Opus 4.7 ↑20 4306.4
- 3
Qwen3.7-Plus ↑16 3964.4
- 4 GPT-5.4 ↑13 3923.1
- 5 GPT-5.5 ↑16 3815.1
- 6 GLM-5V-Turbo ↑10 3138.2
- 7
MiniMax-M3 ↑7 3061.5
- 8
Qwen 3.6 Plus ↑5 3045.5
- 9 Claude Opus 4.6 ↑6 3023.9
- 10
Gemini 3 Flash ↑1 2851.0
Synthetic
23 models we track have a score here.
- 1 Claude Opus 4.8 ↑22 4549.5
- 2 Claude Opus 4.7 ↑20 4501.4
- 3 GPT-5.4 ↑14 4221.1
- 4
Qwen3.7-Plus ↑15 3996.0
- 5 GLM-5V-Turbo ↑11 3859.8
- 6 GPT-5.5 ↑15 3467.8
- 7 Claude Opus 4.6 ↑8 3363.5
- 8 Claude Fable 5 ↑12 3164.2
- 9
Qwen 3.6 Plus ↑4 2956.4
- 10
MiniMax-M3 ↑4 2905.8
Place Naming
23 models we track have a score here.
- 1 GPT-5.5 ↑20 62.5%
- 2 Claude Fable 5 ↑18 62.5%
- 3 GPT-5.4 Pro ↑15 60%
- 4 Claude Opus 4.8 ↑19 60%
- 5 Claude Opus 4.7 ↑17 56.3%
- 6 GPT-5.4 ↑11 50%
- 7 Kimi K2.6 ↓1 50%
- 8 Claude Sonnet 4.6 ↑2 47.5%
- 9
Gemini 3.1 Pro ↓1 47.5%
- 10 GPT-5.5 Pro ↑2 47.5%
Navigation Challenge
23 models we track have a score here.
- 1 GPT-5.4 Pro ↑17 35%
- 2 Claude Fable 5 ↑18 35%
- 3 GPT-5.5 ↑18 32.5%
- 4 Claude Opus 4.7 ↑18 31.3%
- 5 GPT-5.5 Pro ↑7 31.3%
- 6
Gemini 3.1 Pro ↑2 30%
- 7 Claude Opus 4.8 ↑16 26.3%
- 8
Qwen3.7-Plus ↑11 23.8%
- 9
Gemini 3 Flash ↑2 18.8%
- 10 Claude Opus 4.6 ↑5 18.8%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official GeoGuesser Bench 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 GeoGuesser Bench 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
All 9 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access The tasks are private.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (23 entries).
- Pass Models scored 23 models we track.
- Pass Newest score The newest is from June 2026, for Claude Fable 5.
- Note Frontier models None of the top 10 has a score. The strongest tested is Claude Fable 5 (#12).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on huanzhimao.com: 23 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.