Coding

PostTrainBench v1.1 leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, the highest measured score on PostTrainBench v1.1 is 49.3% by Claude Opus 5.5. 175 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	49.3%	measured
2	Gemini 4 Argon	45.3%	measured
3	Claude Sonnet 5.5	44.4%	estimated ± 4.3 pp, medium confidence
4	GPT-6 Astra	44.3%	measured
5	Muse Spark 1.3	43.9%	estimated ± 4.8 pp, high confidence
6	Ember-1	43.7%	estimated ± 4.8 pp, high confidence
7	DeepSeek V4.1 Flash	42.8%	estimated ± 4.8 pp, high confidence
8	Pareto 26.9	42.6%	estimated ± 4.8 pp, high confidence
9	SWE-2	41.6%	estimated ± 4.8 pp, high confidence
10	GPT-6.1 Sol	40.6%	estimated ± 4.8 pp, high confidence
11	MiMo-V2.6-Pro	40.6%	estimated ± 4.8 pp, high confidence
12	Claude Fable 5.1	40.2%	measured
13	Claude Mythos 5	40.1%	estimated ± 4.7 pp, high confidence
14	Pareto 26.10 Preview	38.7%	estimated ± 4.8 pp, high confidence
15	GPT-6 Sol	37.6%	estimated ± 4.8 pp, high confidence
16	MiMo-V2.6-Flash	36.7%	estimated ± 4.8 pp, high confidence
17	Step 5 Preview	36.5%	estimated ± 4.8 pp, high confidence
18	GPT-5.6 Sol	36.2%	measured
19	GLM-5.3	36.2%	estimated ± 2.7 pp, high confidence
20	GPT-6 Luna	35.5%	estimated ± 4.8 pp, high confidence
21	Claude Opus 5	35.0%	measured
22	Sakana Fugu-Ultra	34.9%	estimated ± 4.7 pp, high confidence
23	DeepSeek V4 Pro 0813	34.5%	estimated ± 2.7 pp, high confidence
24	GPT-5.6 Terra	34.4%	estimated ± 2.7 pp, high confidence
25	GLM-5.3-Flash	34.1%	estimated ± 2.7 pp, high confidence
26	Grok 4.6	33.8%	estimated ± 2.7 pp, high confidence
27	Grok 4.7	33.4%	estimated ± 4.3 pp, medium confidence
28	GPT-5.6 Luna	33.1%	estimated ± 3.3 pp, high confidence
29	Claude Fable 5	32.9%	estimated ± 2.7 pp, high confidence
30	Claude Opus 4.8	32.9%	measured
31	Claude Haiku 5.5	32.6%	estimated ± 5.9 pp, low confidence
32	Kimi K3	32.0%	measured
33	GLM-5.2	31.7%	measured
34	Claude Haiku 4.5	30.9%	estimated ± 2.7 pp, high confidence
35	Mistral Large 4	30.4%	estimated ± 7.2 pp, low confidence
36	Ling 3.1 Flash	30.3%	estimated ± 7.2 pp, low confidence
37	DeepSeek V4 Flash 0731	29.8%	estimated ± 2.7 pp, high confidence
38	Hy4 preview	29.3%	estimated ± 4.7 pp, high confidence
39	Qwen3.8-27B	29.1%	estimated ± 2.7 pp, high confidence
40	Beam	29.1%	estimated ± 4.7 pp, high confidence
41	Muse Spark 1.2	29.1%	estimated ± 3.3 pp, high confidence
42	Ornith-1.5-397B	28.9%	estimated ± 4.7 pp, high confidence
43	SWE-1.7	28.8%	estimated ± 5.9 pp, low confidence
44	Claude Opus 4.7	28.6%	measured
45	Claude Opus 4.7 (Adaptive)	28.4%	estimated ± 4.7 pp, high confidence
46	Qwen3.8-Omni-Flash	27.7%	estimated ± 4.7 pp, high confidence
47	Qwen3.8 Max	27.6%	estimated ± 2.7 pp, high confidence
48	GPT-5.5	27.2%	measured
49	Qwen3.8-Flash-Next	27.2%	estimated ± 4.7 pp, high confidence
50	Qwen3.8 Max Preview	27.1%	estimated ± 6.1 pp, medium confidence
51	Ornith-1.0-397B	27.0%	estimated ± 4.7 pp, high confidence
52	dots3-note Preview	26.3%	estimated ± 4.7 pp, high confidence
53	Inkling-Small	26.1%	estimated ± 2.7 pp, high confidence
54	Muse Spark 1.1	26.0%	estimated ± 2.7 pp, high confidence
55	Claude Sonnet 5	25.6%	estimated ± 2.7 pp, high confidence
56	Atria Dawn Preview	25.4%	estimated ± 4.7 pp, high confidence
57	Ornith-1.5-35B-A3B	25.4%	estimated ± 4.7 pp, high confidence
58	Laguna S 2.1	25.3%	estimated ± 4.7 pp, high confidence
59	Composer 2.5	25.2%	estimated ± 3.3 pp, high confidence
60	Sakana Fugu	25.0%	estimated ± 4.7 pp, high confidence
61	Kimi K2.7 Code	24.9%	estimated ± 2.7 pp, high confidence
62	Claude Sonnet 4.6	24.4%	estimated ± 2.7 pp, high confidence
63	Solar Pro 4	24.4%	estimated ± 7.2 pp, low confidence
64	Gemini 3.7 Flash	24.3%	estimated ± 2.7 pp, high confidence
65	Ling 3.0 Flash VL	24.2%	estimated ± 7.2 pp, low confidence
66	Qwen3.7 Plus	24.2%	estimated ± 4.7 pp, medium confidence
67	GLM-5.1	24.0%	estimated ± 2.7 pp, high confidence
68	Claude Opus 4.5	23.9%	estimated ± 4.7 pp, medium confidence
69	Gemini 3.5 Flash-Lite	23.9%	estimated ± 2.7 pp, high confidence
70	Gemini 3.6 Flash	23.8%	estimated ± 2.7 pp, high confidence
71	Gemini 3.8 Flash	23.6%	estimated ± 2.7 pp, high confidence
72	K-EXAONE 2.0	23.5%	estimated ± 7.2 pp, low confidence
73	Grok 4.5	23.5%	measured
74	Gemini 3.5 Flash	23.4%	estimated ± 2.7 pp, high confidence
75	Step 3.7 Flash	23.4%	estimated ± 4.7 pp, medium confidence
76	Inkling	23.4%	estimated ± 2.7 pp, high confidence
77	A.X K2	23.2%	estimated ± 7.2 pp, low confidence
78	GPT-5.3 Codex	23.0%	estimated ± 2.7 pp, high confidence
79	GPT-5.2	23.0%	estimated ± 4.7 pp, medium confidence
80	MiniMax M3	22.8%	estimated ± 2.7 pp, high confidence
81	MiniMax M2.7	22.8%	estimated ± 2.7 pp, high confidence
82	GLM-5	22.8%	estimated ± 4.7 pp, medium confidence
83	DeepSeek V3 0324	22.7%	estimated ± 7.2 pp, low confidence
84	North Mini Code	22.7%	estimated ± 7.2 pp, low confidence
85	Mercury 2.5	22.6%	estimated ± 7.2 pp, low confidence
86	Muse Spark	22.6%	estimated ± 3.3 pp, medium confidence
87	MiMo-V2.5-Pro	22.5%	estimated ± 2.7 pp, high confidence
88	Kimi K2.6	22.1%	estimated ± 2.7 pp, high confidence
89	Gemini 3.1 Pro	22.0%	measured
90	Claude Opus 4.6	21.8%	estimated ± 4.7 pp, medium confidence
91	Qwen 3.6 Max (preview)	21.8%	estimated ± 3.3 pp, medium confidence
92	GPT-5.4 mini	21.8%	estimated ± 2.7 pp, high confidence
93	Gemini 3 Flash	21.7%	estimated ± 2.7 pp, high confidence
94	Solar Pro 3	21.5%	estimated ± 7.2 pp, low confidence
95	MAI-Thinking-1	21.5%	estimated ± 4.7 pp, medium confidence
96	Granite 4.2 3B	21.5%	estimated ± 7.2 pp, low confidence
97	Ling 3.0 Tiny	21.4%	estimated ± 7.2 pp, low confidence
98	Claude Opus 4.6 (Adaptive)	21.1%	estimated ± 10.5 pp, low confidence
99	Muse Glimmer 30B	20.6%	estimated ± 4.7 pp, medium confidence
100	Qwen3.6 Plus	20.6%	estimated ± 2.7 pp, high confidence
101	MiMo-V2.5	20.5%	estimated ± 2.7 pp, high confidence
102	Qwen3.6-27B	20.5%	estimated ± 3.3 pp, medium confidence
103	Qwen3.5 397B	20.4%	estimated ± 4.7 pp, medium confidence
104	Grok 4.20	20.4%	estimated ± 2.7 pp, high confidence
105	Kimi K2.5	20.3%	estimated ± 4.7 pp, medium confidence
106	Ornith-1.0-35B	20.2%	estimated ± 4.7 pp, medium confidence
107	GLM-5 (Reasoning)	19.9%	estimated ± 10.5 pp, low confidence
108	Claude Sonnet 4.5 Thinking	19.9%	estimated ± 10.5 pp, low confidence
109	GPT-5.1-Codex-Max	19.9%	estimated ± 10.5 pp, low confidence
110	Claude Opus 4.5 Thinking	19.9%	estimated ± 10.5 pp, low confidence
111	Qwen3.5 Plus	19.9%	estimated ± 10.5 pp, low confidence
112	MiniMax M2.5	19.9%	estimated ± 10.5 pp, low confidence
113	Gemini 3 Pro	19.9%	estimated ± 10.5 pp, low confidence
114	GPT-5 mini	19.9%	estimated ± 10.5 pp, low confidence
115	GPT-5.1-Codex	19.9%	estimated ± 10.5 pp, low confidence
116	Claude Haiku 4.5 Thinking	19.9%	estimated ± 10.5 pp, low confidence
117	DeepSeek V3.2 (Thinking)	19.9%	estimated ± 10.5 pp, low confidence
118	GLM-4.6	19.9%	estimated ± 10.5 pp, low confidence
119	Grok 4.1 Fast (Reasoning)	19.9%	estimated ± 10.5 pp, low confidence
120	Grok 4 Fast (Reasoning)	19.9%	estimated ± 10.5 pp, low confidence
121	Qwen3 Max	19.9%	estimated ± 10.5 pp, low confidence
122	Grok 4.3	19.8%	estimated ± 2.7 pp, high confidence
123	Qwen3.6-35B-A3B	19.7%	estimated ± 4.7 pp, medium confidence
124	GLM-4.7	19.3%	estimated ± 2.7 pp, high confidence
125	GPT-5.2-Codex	19.3%	estimated ± 2.7 pp, high confidence
126	GPT-5.4	19.0%	measured
127	GPT-5.4 nano	19.0%	estimated ± 2.7 pp, high confidence
128	Mistral Medium 3.5 128B	18.9%	estimated ± 3.3 pp, medium confidence
129	Laguna XS 2.1	18.7%	estimated ± 4.7 pp, medium confidence
130	Ornith-1.5-9B	18.6%	estimated ± 4.7 pp, medium confidence
131	Nemotron 3 Ultra	17.8%	estimated ± 2.7 pp, high confidence
132	Qwen3.7 Max	17.3%	estimated ± 2.7 pp, high confidence
133	Laguna M.1	16.5%	estimated ± 2.7 pp, high confidence
134	Quasar 438B	16.3%	estimated ± 6.1 pp, low confidence
135	Ornith-1.0-9B	16.3%	estimated ± 4.7 pp, medium confidence
136	Ling 3.0 Flash	16.1%	estimated ± 2.7 pp, high confidence
137	Apodex 1.1	16.0%	estimated ± 6.1 pp, low confidence
138	Apodex 1.1 Mini	16.0%	estimated ± 6.1 pp, low confidence
139	Gemini 3.1 Flash-Lite	15.8%	estimated ± 2.7 pp, high confidence
140	Qwen3.5 Flash	15.8%	estimated ± 2.7 pp, high confidence
141	LongCat-Flash-Lite-Sparse	15.3%	estimated ± 4.7 pp, medium confidence
142	Laguna XS.2	15.1%	estimated ± 2.7 pp, high confidence
143	Hy3	14.7%	estimated ± 6.1 pp, low confidence
144	Hy3 Preview	14.7%	estimated ± 6.1 pp, low confidence
145	Gemini 2.5 Pro	14.2%	estimated ± 3.3 pp, medium confidence
146	Granite 4.2 30B	11.9%	estimated ± 4.7 pp, medium confidence
147	Ling 3.0 Flash FP8	11.0%	estimated ± 6.1 pp, low confidence
148	MiMo-V2-Flash	10.8%	estimated ± 6.1 pp, low confidence
149	GPT-5.1	10.7%	estimated ± 6.1 pp, low confidence
150	LLaDA2.2-flash	10.6%	estimated ± 4.7 pp, medium confidence
151	Kimi K2.5 (Reasoning)	10.0%	estimated ± 6.1 pp, low confidence
152	Qwen3.5-122B-A10B	9.8%	estimated ± 6.1 pp, low confidence
153	Gemma 4 31B	9.4%	estimated ± 6.1 pp, low confidence
154	o1	9.0%	estimated ± 6.1 pp, low confidence
155	Gemma 4 26B A4B	8.9%	estimated ± 6.1 pp, low confidence
156	GPT-5 (high)	8.8%	estimated ± 6.1 pp, low confidence
157	Nemotron 3 Super 100B	8.8%	estimated ± 6.1 pp, low confidence
158	o1-preview	8.6%	estimated ± 6.1 pp, low confidence
159	K-Exaone	8.5%	estimated ± 6.1 pp, low confidence
160	Gemma 4 12B	8.5%	estimated ± 6.1 pp, low confidence
161	GPT-OSS 120B	8.4%	estimated ± 6.1 pp, low confidence
162	Command A+	8.4%	estimated ± 6.1 pp, low confidence
163	Nemotron 3.5 Lightning 30B A3B NVFP4	8.4%	estimated ± 6.1 pp, low confidence
164	Mistral Small 4	8.4%	estimated ± 6.1 pp, low confidence
165	Mistral Small 4 (Reasoning)	8.4%	estimated ± 6.1 pp, low confidence
166	Trinity-Large-Preview	8.4%	estimated ± 6.1 pp, low confidence
167	Trinity-Large-Thinking	8.4%	estimated ± 6.1 pp, low confidence
168	Ling 2.6 Flash	8.4%	estimated ± 6.1 pp, low confidence
169	Gemini 1.5 Pro	8.3%	estimated ± 6.1 pp, low confidence
170	DeepSeek V3	8.3%	estimated ± 6.1 pp, low confidence
171	GPT-4 Turbo	8.3%	estimated ± 6.1 pp, low confidence
172	GPT-OSS 20B	8.3%	estimated ± 6.1 pp, low confidence
173	GPT-4.1 mini	8.3%	estimated ± 6.1 pp, low confidence
174	Mistral Large 3	8.3%	estimated ± 6.1 pp, low confidence
175	Claude 3 Opus	8.3%	estimated ± 6.1 pp, low confidence
176	Llama 4 Maverick	8.3%	estimated ± 6.1 pp, low confidence
177	Celeris-1	8.3%	estimated ± 6.1 pp, low confidence
178	Nemotron 3 Nano 30B	8.3%	estimated ± 6.1 pp, low confidence
179	Nemotron 3 Nano Omni 30B A3B	8.3%	estimated ± 6.1 pp, low confidence
180	Ultravox v0.6 Llama 3.3 70B	8.3%	estimated ± 6.1 pp, low confidence
181	GPT-4o mini	8.3%	estimated ± 6.1 pp, low confidence
182	GPT-4.1 nano	8.3%	estimated ± 6.1 pp, low confidence
183	Gemma 3 27B	8.3%	estimated ± 6.1 pp, low confidence
184	Gemma 4 E4B	8.3%	estimated ± 6.1 pp, low confidence
185	Gemma 4 E2B	8.3%	estimated ± 6.1 pp, low confidence
186	LFM2.5-2.6B	8.3%	estimated ± 6.1 pp, low confidence
187	Llama 4 Scout	8.3%	estimated ± 6.1 pp, low confidence
188	Granite 4.2 8B	6.3%	estimated ± 4.7 pp, medium confidence
189	MiniCPM5-2B	4.6%	estimated ± 4.7 pp, medium confidence

PostTrainBench v1.1 leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General