Knowledge & reasoning

BioBasicsBench leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap
License: https://creativecommons.org/licenses/by-nc/4.0/

As of 2026-11-07, the highest measured score on BioBasicsBench is 67.0% by GPT-6 Astra. 242 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Opus 5.5	69.0%	estimated ± 2.8 pp, low confidence
2	Claude Fable 5.1	68.9%	estimated ± 2.8 pp, low confidence
3	Gemini 4 Argon	68.6%	estimated ± 2.8 pp, low confidence
4	Claude Fable 5	67.7%	estimated ± 2.8 pp, low confidence
5	Claude Sonnet 5.5	67.2%	estimated ± 2.8 pp, low confidence
6	Claude Opus 5	67.1%	estimated ± 2.8 pp, low confidence
7	GPT-6 Astra	67.0%	measured
8	Muse Wonder Full	64.8%	estimated ± 2.8 pp, medium confidence
9	GPT-6.1 Sol	62.6%	measured
10	Sakana Fugu	57.4%	estimated ± 2.2 pp, medium confidence
11	Sakana Fugu-Ultra	57.4%	estimated ± 2.2 pp, medium confidence
12	Muse Wonder	48.7%	estimated ± 2.8 pp, medium confidence
13	GPT-5.6 Sol	45.6%	measured
14	MiMo-V2.6-Pro	44.9%	estimated ± 2.8 pp, medium confidence
15	Claude Opus 4.7 (Adaptive)	41.9%	estimated ± 2.2 pp, medium confidence
16	Kimi K3.1 Pro	41.9%	estimated ± 2.2 pp, medium confidence
17	Muse Spark 1.3	41.8%	estimated ± 2.8 pp, medium confidence
18	Claude Mythos 5	41.0%	estimated ± 2.2 pp, medium confidence
19	MiniMax M3.1 Flash	39.3%	estimated ± 2.8 pp, medium confidence
20	Gemini 3.7 Flash	39.0%	estimated ± 2.8 pp, medium confidence
21	GPT-6 Sol	39.0%	estimated ± 2.8 pp, medium confidence
22	Gemini 3.8 Flash	38.7%	estimated ± 2.8 pp, medium confidence
23	GPT-6 Luna	37.6%	measured
24	Claude Opus 4.8	37.2%	estimated ± 2.2 pp, medium confidence
25	Gemini 3.1 Pro	36.9%	estimated ± 2.8 pp, medium confidence
26	Grok 4.20	36.8%	estimated ± 3.5 pp, low confidence
27	Kimi K3	36.5%	estimated ± 2.2 pp, medium confidence
28	Step 5 Preview	36.1%	estimated ± 2.8 pp, medium confidence
29	Gemini 4 Carbon	36.0%	estimated ± 3.8 pp, medium confidence
30	GPT-6.1 Terra	35.7%	estimated ± 2.8 pp, medium confidence
31	Muse Spark 1.1	35.7%	estimated ± 2.8 pp, medium confidence
32	GPT-5.5	35.4%	measured
33	Muse Spark 1.2	35.1%	estimated ± 2.8 pp, medium confidence
34	Claude Haiku 5.5	34.5%	estimated ± 2.8 pp, medium confidence
35	GPT-5.5 Pro	34.3%	estimated ± 5.1 pp, low confidence
36	GPT-5.4 Pro	34.3%	estimated ± 5.1 pp, low confidence
37	dots3-note Preview	34.3%	estimated ± 5.1 pp, low confidence
38	Claude Sonnet 4.5	34.3%	estimated ± 5.1 pp, low confidence
39	Claude Sonnet 4.5 Thinking	34.3%	estimated ± 5.1 pp, low confidence
40	Gemini 3 Pro Deep Think	34.3%	estimated ± 5.1 pp, low confidence
41	Muse Spark Open	34.3%	estimated ± 2.8 pp, medium confidence
42	Grok 4.7	34.1%	estimated ± 2.8 pp, medium confidence
43	Qwen3.8 Max Preview	34.1%	estimated ± 2.8 pp, medium confidence
44	Grok 4.6	34.1%	estimated ± 2.8 pp, medium confidence
45	Grok 4.5	34.1%	estimated ± 2.8 pp, medium confidence
46	GPT-5.3 Codex	34.1%	estimated ± 2.8 pp, medium confidence
47	Gemini 3 Flash	34.1%	estimated ± 3.8 pp, medium confidence
48	Claude 3.7 Sonnet	34.1%	estimated ± 3.8 pp, low confidence
49	Claude 3.7 Sonnet Thinking	34.1%	estimated ± 3.8 pp, low confidence
50	Claude 3 Haiku	34.1%	estimated ± 3.8 pp, low confidence
51	Claude 4 Sonnet	34.1%	estimated ± 3.8 pp, low confidence
52	Claude Opus 4.5	34.1%	estimated ± 3.8 pp, medium confidence
53	DeepSeek-R1	34.1%	estimated ± 3.8 pp, low confidence
54	DeepSeek V3	34.1%	estimated ± 3.8 pp, low confidence
55	DeepSeek V3 0324	34.1%	estimated ± 3.8 pp, low confidence
56	DeepSeek V3.1	34.1%	estimated ± 3.8 pp, low confidence
57	DeepSeek V3.1 (Reasoning)	34.1%	estimated ± 3.8 pp, low confidence
58	DeepSeek V3.2	34.1%	estimated ± 3.8 pp, low confidence
59	Gemini 2.0 Flash	34.1%	estimated ± 3.8 pp, low confidence
60	Gemini 2.5 Flash	34.1%	estimated ± 3.8 pp, low confidence
61	Gemini 3.5 Flash-Lite	34.1%	estimated ± 3.8 pp, low confidence
62	GLM-4.6	34.1%	estimated ± 3.8 pp, low confidence
63	GLM-5V-Turbo	34.1%	estimated ± 3.8 pp, low confidence
64	GPT-4.1	34.1%	estimated ± 3.8 pp, low confidence
65	GPT-4.1 mini	34.1%	estimated ± 3.8 pp, low confidence
66	GPT-4o	34.1%	estimated ± 3.8 pp, low confidence
67	Grok 4.1 Fast (Reasoning)	34.1%	estimated ± 3.8 pp, low confidence
68	Grok 4 Fast (Reasoning)	34.1%	estimated ± 3.8 pp, low confidence
69	Grok Code Fast 1	34.1%	estimated ± 3.8 pp, low confidence
70	Kimi K2	34.1%	estimated ± 3.8 pp, low confidence
71	Ling 3.0 Tiny	34.1%	estimated ± 3.8 pp, low confidence
72	Magistral Medium	34.1%	estimated ± 3.8 pp, low confidence
73	Magistral Small	34.1%	estimated ± 3.8 pp, low confidence
74	Mercury 2.5	34.1%	estimated ± 3.8 pp, low confidence
75	Mistral Large 2	34.1%	estimated ± 3.8 pp, low confidence
76	Mistral Large 3	34.1%	estimated ± 3.8 pp, low confidence
77	Mistral Large 4	34.1%	estimated ± 3.8 pp, low confidence
78	Mistral Medium 3	34.1%	estimated ± 3.8 pp, low confidence
79	Mistral Medium 3.5 128B	34.1%	estimated ± 3.8 pp, low confidence
80	Mistral Small 4	34.1%	estimated ± 3.8 pp, low confidence
81	Mistral Small 4 (Reasoning)	34.1%	estimated ± 3.8 pp, low confidence
82	Nemotron 3.5 Bolt 74B-A9B BF16	34.1%	estimated ± 3.8 pp, low confidence
83	North Mini Code	34.1%	estimated ± 3.8 pp, low confidence
84	o1	34.1%	estimated ± 3.8 pp, low confidence
85	o4-mini	34.1%	estimated ± 3.8 pp, low confidence
86	Qwen3 Max	34.1%	estimated ± 3.8 pp, low confidence
87	Solar Pro 3	34.1%	estimated ± 3.8 pp, low confidence
88	GLM-5.3	34.0%	estimated ± 2.8 pp, medium confidence
89	Claude Sonnet 5	34.0%	estimated ± 2.8 pp, medium confidence
90	Gemini 3.6 Flash	34.0%	estimated ± 2.8 pp, medium confidence
91	Muse Spark	34.0%	estimated ± 2.8 pp, medium confidence
92	Ling 3.0 Flash Sante	33.9%	estimated ± 6.7 pp, low confidence
93	Sage-A1.5	33.9%	estimated ± 10.7 pp, low confidence
94	Beam 1.0 501B-A23B FP8	33.9%	estimated ± 10.7 pp, low confidence
95	Beam 1.0 501B-A23B NVFP4	33.9%	estimated ± 10.7 pp, low confidence
96	Celeris-1	33.9%	estimated ± 10.7 pp, low confidence
97	Exaone 4.0 1.2B	33.9%	estimated ± 10.7 pp, low confidence
98	Exaone 4.0 32B	33.9%	estimated ± 10.7 pp, low confidence
99	Gemma 4 12B	33.9%	estimated ± 10.7 pp, low confidence
100	Gemma 4 26B A4B	33.9%	estimated ± 10.7 pp, low confidence
101	Gemma 4 E2B	33.9%	estimated ± 10.7 pp, low confidence
102	Gemma 4 E4B	33.9%	estimated ± 10.7 pp, low confidence
103	GPT-OSS 120B	33.9%	estimated ± 10.7 pp, low confidence
104	GPT-OSS 20B	33.9%	estimated ± 10.7 pp, low confidence
105	Granite-4.0-350M	33.9%	estimated ± 10.7 pp, low confidence
106	Granite-4.0-H-1B	33.9%	estimated ± 10.7 pp, low confidence
107	Granite-4.0-H-350M	33.9%	estimated ± 10.7 pp, low confidence
108	Granite 4.2 30B	33.9%	estimated ± 10.7 pp, low confidence
109	K-EXAONE 2.0	33.9%	estimated ± 10.7 pp, low confidence
110	LFM2.5-2.6B	33.9%	estimated ± 10.7 pp, low confidence
111	LFM2.5-8B-A1B	33.9%	estimated ± 10.7 pp, low confidence
112	LFM2.5-VL-1.6B-Extract	33.9%	estimated ± 10.7 pp, low confidence
113	Llama 3.1 405B	33.9%	estimated ± 10.7 pp, low confidence
114	Llama 3.1 70B	33.9%	estimated ± 10.7 pp, low confidence
115	Llama 3.3 70B	33.9%	estimated ± 10.7 pp, low confidence
116	Llama 4 Maverick	33.9%	estimated ± 10.7 pp, low confidence
117	Llama 4 Scout	33.9%	estimated ± 10.7 pp, low confidence
118	MiniCPM5-2B	33.9%	estimated ± 10.7 pp, low confidence
119	Nemotron 3 Nano 30B	33.9%	estimated ± 10.7 pp, low confidence
120	Nemotron 3 Nano Omni 30B A3B	33.9%	estimated ± 10.7 pp, low confidence
121	Nemotron Ultra 253B	33.9%	estimated ± 10.7 pp, low confidence
122	o3	33.9%	estimated ± 10.7 pp, low confidence
123	Phi-4	33.9%	estimated ± 10.7 pp, low confidence
124	Phi-4-reasoning	33.9%	estimated ± 10.7 pp, low confidence
125	Quasar 438B	33.9%	estimated ± 10.7 pp, low confidence
126	Qwen3.5-4B	33.9%	estimated ± 10.7 pp, low confidence
127	Qwen3-Omni-30B-A3B-Instruct	33.9%	estimated ± 10.7 pp, low confidence
128	Qwen3-Omni-30B-A3B-Thinking	33.9%	estimated ± 10.7 pp, low confidence
129	Sarvam 105B	33.9%	estimated ± 10.7 pp, low confidence
130	Sarvam 30B	33.9%	estimated ± 10.7 pp, low confidence
131	Trinity-Large-Preview	33.9%	estimated ± 10.7 pp, low confidence
132	Trinity-Large-Thinking	33.9%	estimated ± 10.7 pp, low confidence
133	Ultravox v0.6 Llama 3.3 70B	33.9%	estimated ± 10.7 pp, low confidence
134	Claude Opus 4.6 (Adaptive)	33.9%	estimated ± 2.8 pp, medium confidence
135	Gemini 3 Pro	33.9%	estimated ± 2.8 pp, medium confidence
136	Ling 3.1 Flash	33.9%	estimated ± 2.8 pp, medium confidence
137	MiniMax M3	33.9%	estimated ± 2.8 pp, medium confidence
138	DeepSeek V4 Flash 0731	33.9%	estimated ± 2.8 pp, medium confidence
139	GPT-5.2-Codex	33.9%	estimated ± 2.8 pp, low confidence
140	MiMo-V2.5-Pro	33.9%	estimated ± 2.8 pp, low confidence
141	MiMo-V2.6-Flash	33.9%	estimated ± 2.8 pp, low confidence
142	Kimi K2.7 Code	33.9%	estimated ± 2.8 pp, low confidence
143	Beam 1.0 501B-A23B	33.9%	estimated ± 2.8 pp, low confidence
144	Apodex 1.1	33.9%	estimated ± 2.8 pp, low confidence
145	Apodex 1.1 Mini	33.9%	estimated ± 2.8 pp, low confidence
146	Hy3	33.9%	estimated ± 2.8 pp, low confidence
147	Hy3 Preview	33.9%	estimated ± 2.8 pp, low confidence
148	Claude Opus 4.7	33.9%	estimated ± 2.8 pp, low confidence
149	Inkling	33.9%	estimated ± 2.8 pp, low confidence
150	Qwen 3.6 Max (preview)	33.9%	estimated ± 2.8 pp, low confidence
151	Kimi K2.5	33.9%	estimated ± 2.8 pp, low confidence
152	Kimi K2.5 (Reasoning)	33.9%	estimated ± 2.8 pp, low confidence
153	Nemotron 3.5 Super VL 120B A12B	33.9%	estimated ± 2.8 pp, low confidence
154	Nemotron 3.5 Super VL 120B A12B NVFP4	33.9%	estimated ± 2.8 pp, low confidence
155	MiMo-V2-Pro	33.9%	estimated ± 2.8 pp, low confidence
156	Claude Opus 4.5 Thinking	33.9%	estimated ± 2.8 pp, low confidence
157	GLM-5.1	33.9%	estimated ± 2.8 pp, low confidence
158	Solar Pro 4	33.9%	estimated ± 2.8 pp, low confidence
159	A.X K2	33.9%	estimated ± 2.8 pp, low confidence
160	MiniMax M2.7	33.9%	estimated ± 2.8 pp, low confidence
161	GLM-5	33.9%	estimated ± 2.8 pp, low confidence
162	Qwen3.5 397B	33.9%	estimated ± 2.8 pp, low confidence
163	Qwen3.5 397B (Reasoning)	33.9%	estimated ± 2.8 pp, low confidence
164	GPT-5.1	33.9%	estimated ± 2.8 pp, low confidence
165	GPT-5 (high)	33.9%	estimated ± 2.8 pp, low confidence
166	Nemotron 3 Ultra	33.9%	estimated ± 2.8 pp, low confidence
167	GPT-5.4 nano	33.9%	estimated ± 2.8 pp, low confidence
168	GPT-5.4 mini	33.9%	estimated ± 2.8 pp, low confidence
169	GLM-5-Turbo	33.9%	estimated ± 2.8 pp, low confidence
170	GLM-4.7	33.9%	estimated ± 2.8 pp, low confidence
171	Grok 4	33.9%	estimated ± 2.8 pp, low confidence
172	GPT-5.1-Codex	33.9%	estimated ± 2.8 pp, low confidence
173	GPT-5.1-Codex-Max	33.9%	estimated ± 2.8 pp, low confidence
174	GPT-5 (medium)	33.9%	estimated ± 2.8 pp, low confidence
175	Qwen3.5-122B-A10B	33.9%	estimated ± 2.8 pp, low confidence
176	Nemotron 3.5 Bolt 74B-A9B	33.9%	estimated ± 2.8 pp, low confidence
177	Qwen3.5-27B	33.9%	estimated ± 2.8 pp, low confidence
178	Ling 3.0 Flash	33.9%	estimated ± 2.8 pp, low confidence
179	Gemma 4 31B	33.9%	estimated ± 2.8 pp, low confidence
180	Qwen3.6-27B	33.9%	estimated ± 2.8 pp, low confidence
181	Gemini 2.5 Pro	33.9%	estimated ± 2.8 pp, low confidence
182	Qwen3.6-35B-A3B	33.9%	estimated ± 2.8 pp, low confidence
183	MiMo-V2-Omni	33.9%	estimated ± 2.8 pp, low confidence
184	Ling 3.0 Flash VL	33.9%	estimated ± 2.8 pp, low confidence
185	Muse Glimmer 30B	33.9%	estimated ± 2.8 pp, low confidence
186	Step 3.7 Flash	33.9%	estimated ± 2.8 pp, low confidence
187	Qwen3.5-35B-A3B	33.9%	estimated ± 2.8 pp, low confidence
188	Inkling 2 Small	33.4%	estimated ± 2.2 pp, medium confidence
189	GPT-5.4	33.2%	measured
190	GPT-5.6 Terra	32.8%	estimated ± 2.2 pp, medium confidence
191	Ornith-1.5-397B	32.3%	estimated ± 2.2 pp, medium confidence
192	Qwen3.8 Max	31.2%	estimated ± 2.2 pp, medium confidence
193	GPT-5.2	31.2%	measured
194	Qwen3.7 Max	30.2%	estimated ± 2.2 pp, medium confidence
195	GPT-5.6 Luna	29.8%	estimated ± 2.2 pp, low confidence
196	Hy4 preview	29.8%	estimated ± 2.2 pp, low confidence
197	Gemini 3.5 Flash	29.3%	estimated ± 2.2 pp, low confidence
198	Qwen3.8-Flash-Next	27.2%	estimated ± 2.2 pp, low confidence
199	Claude 4.1 Opus Thinking	26.6%	estimated ± 10.2 pp, low confidence
200	Command A+	26.3%	estimated ± 10.2 pp, low confidence
201	Claude Opus 4.6	25.7%	estimated ± 2.2 pp, low confidence
202	o3-pro	25.6%	estimated ± 10.2 pp, low confidence
203	GLM-5.2	25.4%	estimated ± 2.2 pp, low confidence
204	Qwen3.8-Omni-Flash	24.7%	estimated ± 2.2 pp, low confidence
205	DeepSeek V4.1 Flash 1017	24.4%	estimated ± 2.2 pp, low confidence
206	Claude 4 Opus Thinking	24.3%	estimated ± 10.2 pp, low confidence
207	Kimi K2.6	23.1%	estimated ± 2.2 pp, low confidence
208	Qwen3.6 Plus	22.8%	estimated ± 2.2 pp, low confidence
209	Qwen3.7 Plus	22.6%	estimated ± 2.2 pp, low confidence
210	Claude 4.1 Opus	22.2%	estimated ± 10.2 pp, low confidence
211	DeepSeek V4 Pro 0813	22.0%	estimated ± 2.2 pp, low confidence
212	Grok 4.3	22.0%	estimated ± 2.2 pp, low confidence
213	Claude Sonnet 4.6	21.5%	estimated ± 2.2 pp, low confidence
214	Interfaze Beta	21.5%	estimated ± 2.2 pp, low confidence
215	Inkling-Small	20.5%	estimated ± 2.2 pp, low confidence
216	Claude 4 Opus	20.1%	estimated ± 10.2 pp, low confidence
217	Ornith-1.5-35B-A3B	19.8%	estimated ± 2.2 pp, low confidence
218	Qwen3.8-27B	19.8%	estimated ± 2.2 pp, low confidence
219	MiMo-V2-Flash	19.5%	estimated ± 10.2 pp, low confidence
220	K-Exaone	17.6%	estimated ± 10.2 pp, low confidence
221	Ling 2.6 Flash	17.2%	estimated ± 10.2 pp, low confidence
222	o3-mini	15.4%	estimated ± 10.2 pp, low confidence
223	o1-pro	15.4%	estimated ± 10.2 pp, low confidence
224	o1-preview	14.2%	estimated ± 10.2 pp, low confidence
225	Grok 4.1 Fast	14.1%	estimated ± 10.2 pp, low confidence
226	GLM-4.5-Air	13.9%	estimated ± 10.2 pp, low confidence
227	o1-mini	12.3%	estimated ± 10.2 pp, low confidence
228	GPT-4.5	12.1%	estimated ± 10.2 pp, low confidence
229	QwQ-32B	12.0%	estimated ± 10.2 pp, low confidence
230	Claude 3.5 Haiku	11.3%	estimated ± 10.2 pp, low confidence
231	Claude 3 Opus	11.1%	estimated ± 10.2 pp, low confidence
232	DeepSeek R1 Distill Qwen 32B	10.7%	estimated ± 10.2 pp, low confidence
233	Gemini 1.5 Pro	10.1%	estimated ± 10.2 pp, low confidence
234	GPT-4.1 nano	10.0%	estimated ± 10.2 pp, low confidence
235	Devstral Small	9.7%	estimated ± 10.2 pp, low confidence
236	Mistral Small 3.1	9.1%	estimated ± 10.2 pp, low confidence
237	Gemini 1.5 Flash	9.1%	estimated ± 10.2 pp, low confidence
238	GPT-4 Turbo	9.0%	estimated ± 10.2 pp, low confidence
239	Solar Pro 2	9.0%	estimated ± 10.2 pp, low confidence
240	Nova Pro	8.9%	estimated ± 10.2 pp, low confidence
241	Qwen2.5 Coder 32B Instruct	8.7%	estimated ± 10.2 pp, low confidence
242	Gemini 2.5 Flash-Lite	8.6%	estimated ± 10.2 pp, low confidence
243	GPT-4o mini	8.6%	estimated ± 10.2 pp, low confidence
244	DeepSeek-V2.5	8.5%	estimated ± 10.2 pp, low confidence
245	Nemotron 3.5 Lightning 30B A3B NVFP4	7.2%	estimated ± 10.2 pp, low confidence
246	Gemini 1.0 Pro	6.9%	estimated ± 10.2 pp, low confidence
247	Granite 4.2 8B	6.6%	estimated ± 10.2 pp, low confidence
248	Gemma 3 27B	6.3%	estimated ± 10.2 pp, low confidence
249	Granite 4.2 3B	6.1%	estimated ± 10.2 pp, low confidence

BioBasicsBench leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General