WildlifeBench Wildlife Knowledge V1
Models answer 185 questions about wild species without search or tools. The overall score weights ten tasks equally, including taxonomy, conservation status, and geographic range.
The maker's leaderboard scores seven general-purpose models from seven developers. The site publishes practice examples and aggregate scores. The scored questions remain private.
- Made by
- WildlifeBench
- Released
- November 2026
- Kind
- Science questions
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- Public, plus a private test set
- Measures
- Capability
- Subject
- Biological sciences
- Language
- English
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- A host
- In views
- Biology and medicine
- Strongest model tested
- GPT-6.1 Sol #6 on the capability index
- Last model added
- October 2026
- Scores refreshed
- November 20th, 2026
Muse Spark 1.3 leads with 84.7%
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.
- 1
Muse Spark 1.3 Zero-shot, single pass 84.7%
- 2 GPT-6.1 Sol Zero-shot, single pass 83.5%
- 3
Gemini 4 Argon Zero-shot, single pass 80.5%
- 4 Grok 4.7 Zero-shot, single pass 72.9%
- 5
Qwen3.8-2.4T-A95B Zero-shot, single pass 63.5%
- 6
Mistral Large 4 Zero-shot, single pass 60.8%
- 7 Claude Haiku 5.5 Zero-shot, single pass 57.4%
Source: Official WildlifeBench Wildlife Knowledge V1 leaderboard
| # | Model | Score ± | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 84.7% 81.3 to 88 | Zero-shot, single pass | Sep 2026 | Official leaderboard |
| 2 | GPT-6.1 Sol | 83.5% ±3.6 | Zero-shot, single pass | Sep 2026 | Official leaderboard |
| 3 | | 80.5% ±3.5 | Zero-shot, single pass | Sep 2026 | Official leaderboard |
| 4 | Grok 4.7 | 72.9% ±4.9 | Zero-shot, single pass | Sep 2026 | Official leaderboard |
| 5 | | 63.5% ±4.6 | Zero-shot, single pass | Aug 2026 | Official leaderboard |
| 6 | | 60.8% ±4.5 | Zero-shot, single pass | Oct 2026 | Official leaderboard |
| 7 | Claude Haiku 5.5 | 57.4% ±4.5 | Zero-shot, single pass | Oct 2026 | Official leaderboard |
Rankings by part
We read 15 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Taxonomy, Existence and Red List. Muse Spark 1.3 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.
Taxonomy
7 models we track have a score here.
- 1
Muse Spark 1.3 100%
- 2 GPT-6.1 Sol 95%
- 3
Gemini 4 Argon 95%
- 4 Grok 4.7 90%
- 5
Qwen3.8-2.4T-A95B 75%
- 6
Mistral Large 4 75%
- 7 Claude Haiku 5.5 75%
Existence
7 models we track have a score here.
- 1
Muse Spark 1.3 94.3%
- 2
Gemini 4 Argon ↑1 93.7%
- 3 Grok 4.7 ↑1 89.7%
- 4 GPT-6.1 Sol ↓2 83.4%
- 5 Claude Haiku 5.5 ↑2 74.3%
- 6
Mistral Large 4 68.4%
- 7
Qwen3.8-2.4T-A95B ↓2 57.5%
Red List
7 models we track have a score here.
- 1
Muse Spark 1.3 64.9%
- 2 GPT-6.1 Sol 64.9%
- 3
Gemini 4 Argon 64.9%
- 4
Mistral Large 4 ↑2 59.5%
- 5
Qwen3.8-2.4T-A95B 54.1%
- 6 Grok 4.7 ↓2 54.1%
- 7 Claude Haiku 5.5 43.2%
Native range
7 models we track have a score here.
- 1 GPT-6.1 Sol ↑1 69.5%
- 2
Muse Spark 1.3 ↓1 66.6%
- 3
Gemini 4 Argon 59.4%
- 4 Grok 4.7 49%
- 5
Mistral Large 4 ↑1 42.8%
- 6
Qwen3.8-2.4T-A95B ↓1 42.6%
- 7 Claude Haiku 5.5 31.8%
Introduced range
7 models we track have a score here.
- 1 GPT-6.1 Sol ↑1 73.8%
- 2
Muse Spark 1.3 ↓1 62.1%
- 3
Mistral Large 4 ↑3 58.7%
- 4
Qwen3.8-2.4T-A95B ↑1 44.9%
- 5 Grok 4.7 ↓1 36.4%
- 6
Gemini 4 Argon ↓3 36.4%
- 7 Claude Haiku 5.5 34.7%
Checklists
7 models we track have a score here.
- 1
Gemini 4 Argon ↑2 97.9%
- 2
Muse Spark 1.3 ↓1 96.5%
- 3 GPT-6.1 Sol ↓1 96%
- 4 Grok 4.7 92%
- 5
Mistral Large 4 ↑1 87.6%
- 6 Claude Haiku 5.5 ↑1 84.5%
- 7
Qwen3.8-2.4T-A95B ↓2 84.5%
Common names
7 models we track have a score here.
- 1
Muse Spark 1.3 100%
- 2 Grok 4.7 ↑2 100%
- 3 GPT-6.1 Sol ↓1 100%
- 4
Gemini 4 Argon ↓1 100%
- 5
Qwen3.8-2.4T-A95B 85%
- 6 Claude Haiku 5.5 ↑1 75%
- 7
Mistral Large 4 ↓1 65%
Authority
7 models we track have a score here.
- 1
Gemini 4 Argon ↑2 100%
- 2 GPT-6.1 Sol 95%
- 3
Muse Spark 1.3 ↓2 90%
- 4 Grok 4.7 80%
- 5
Qwen3.8-2.4T-A95B 50%
- 6
Mistral Large 4 25%
- 7 Claude Haiku 5.5 25%
Invasive species
7 models we track have a score here.
- 1
Muse Spark 1.3 93%
- 2
Gemini 4 Argon ↑1 92.5%
- 3 GPT-6.1 Sol ↓1 92%
- 4 Grok 4.7 87.5%
- 5
Mistral Large 4 ↑1 85.5%
- 6 Claude Haiku 5.5 ↑1 85.5%
- 7
Qwen3.8-2.4T-A95B ↓2 81.5%
Nomenclature
7 models we track have a score here.
- 1
Muse Spark 1.3 80%
- 2 GPT-6.1 Sol 65%
- 3
Gemini 4 Argon 65%
- 4
Qwen3.8-2.4T-A95B ↑1 60%
- 5 Grok 4.7 ↓1 50%
- 6 Claude Haiku 5.5 ↑1 45%
- 7
Mistral Large 4 ↓1 40%
Invented names rejected
7 models we track have a score here.
- 1 Grok 4.7 ↑3 100%
- 2
Gemini 4 Argon ↑1 100%
- 3
Muse Spark 1.3 ↓2 95%
- 4 GPT-6.1 Sol ↓2 70%
- 5 Claude Haiku 5.5 ↑2 55%
- 6
Mistral Large 4 40%
- 7
Qwen3.8-2.4T-A95B ↓2 15%
Real names accepted
7 models we track have a score here.
- 1
Qwen3.8-2.4T-A95B ↑4 100%
- 2 GPT-6.1 Sol 96.8%
- 3
Mistral Large 4 ↑3 96.8%
- 4
Muse Spark 1.3 ↓3 93.7%
- 5 Claude Haiku 5.5 ↑2 93.7%
- 6
Gemini 4 Argon ↓3 87.3%
- 7 Grok 4.7 ↓3 79.4%
Mean accuracy across categories
7 models we track have a score here.
- 1
Gemini 4 Argon ↑2 60.9%
- 2
Muse Spark 1.3 ↓1 60.1%
- 3 GPT-6.1 Sol ↓1 57.7%
- 4 Grok 4.7 49.4%
- 5
Mistral Large 4 ↑1 48%
- 6
Qwen3.8-2.4T-A95B ↓1 43.7%
- 7 Claude Haiku 5.5 32.1%
Standard English name
7 models we track have a score here.
- 1 GPT-6.1 Sol ↑1 80%
- 2 Grok 4.7 ↑2 75%
- 3
Gemini 4 Argon 75%
- 4
Muse Spark 1.3 ↓3 70%
- 5 Claude Haiku 5.5 ↑2 55%
- 6
Qwen3.8-2.4T-A95B ↓1 50%
- 7
Mistral Large 4 ↓1 45%
Recorded as alien, invasive or not
7 models we track have a score here.
- 1
Gemini 4 Argon ↑2 98.9%
- 2 GPT-6.1 Sol 97.9%
- 3
Muse Spark 1.3 ↓2 97.5%
- 4 Claude Haiku 5.5 ↑3 95.4%
- 5 Grok 4.7 ↓1 95%
- 6
Qwen3.8-2.4T-A95B ↓1 94.5%
- 7
Mistral Large 4 ↓1 93.4%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official WildlifeBench Wildlife Knowledge V1 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 WildlifeBench Wildlife Knowledge V1 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
- 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 Public tasks, plus a private test set.
- Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
- Pass Independent runs Scores from the official leaderboard (7 entries).
- Pass Models scored 7 models we track.
- Pass Newest score The newest is from October 2026, for Claude Haiku 5.5.
- Note Frontier models Only 2 of the top 10 have a score. The strongest tested is GPT-6.1 Sol (#6).
- Pass Headroom The best is 84.7%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on wildlifebench.org: 7 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.