The leaders, at a glance

Agentic AI model ranking

Compare AI models and effort settings for agentic work using sourced task evaluations and visible evidence coverage.

See the full ranking ↓How configurations are ranked

Agentic leaders

Experimental comparison score · 0–100

For the selected profile and sources. Scores are experimental; small gaps may not be meaningful. How scoring works · Effort evidence collected 2026-09-12

451unique modelsEach model counted once across effort levels 492model configurationsDocumented model + effort settings, including unranked

109aggregated sourcesOriginal reports and benchmark leaderboards

Explore all 109 sources Independent evaluations and lab reports

Distinct source pages with published results in the catalog. Multiple results and page sections count as one source.

Every ranked row is a measured configuration. The default view shows the highest-scoring supported configuration per model. In Value, the budget is applied first. Select All efforts for the full comparison. How effort ranking works.

Customize ranking

Full ranking

127 entries

Custom capability ranking of measured model configurations. Scores are experimental; missing results remain unknown.
Rank	Model	Score	Capability coverage	analyst-agent	briefcase	aime	apex-agents	aa-lcr	automationbench	critpt	enterprise-ops	gdp-pdf	gdpval	gmmlu	gpqa	harvey	hle	ifbench	itbench	livecodebench	math500	mmlu-pro	mmmu-pro	omniscience	scicode	tau-bench	terminal-bench	arc-agi	deepswe	swe-bench-pro	arc-agi-3	frontiercode	enigma-eval	swe-atlas-qna	swe-atlas-refactoring	swe-atlas-test-writing	livebench-math	livebench-coding	livebench-language	livebench-data	multi-swe-bench	livebench-reasoning	livebench-instructions	swe-rebench	epoch-game-puzzles	frontiermath	simpleqa	vals-finance-agent	vals-code-migration	vals-legal-research	vals-excel-modeling	vibe-code	chartography	osworld	charxiv	swe-bench-multilingual	deepsearchqa	officeqa	mcp-atlas	toolathlon	mrcr	browsecomp
101	Qwen3.5-0.8B · Non-reasoningQwen Measured effort: Non-reasoning	5.2	Eligible · 100%12 published · 2 families · 1 selected capabilities	—	—	—	—	14.2197/197 peersIndependent evidence	—	25.1200/200 peersIndependent evidence	—	—	2.1123/123 peersIndependent evidence	—	1.8212/212 peersIndependent evidence	—	30.3208/208 peersIndependent evidence	4.0178/178 peersIndependent evidence	—	—	—	—	1.5114/114 peersIndependent evidence	1.3198/198 peersIndependent evidence	—	61.9171/171 peersIndependent evidence	3.2201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
102	Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 · ReasoningNVIDIA Measured effort: Reasoning	5.0	Eligible · 100%12 published · 2 families · 1 selected capabilities	—	—	—	—	41.1197/197 peersIndependent evidence	—	25.3200/200 peersIndependent evidence	—	—	21.9123/123 peersIndependent evidence	—	15.5212/212 peersIndependent evidence	—	26.3208/208 peersIndependent evidence	75.7178/178 peersIndependent evidence	—	—	—	—	25.9114/114 peersIndependent evidence	27.9198/198 peersIndependent evidence	—	53.9171/171 peersIndependent evidence	34.4201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
103	Magistral Medium 1.2 · ReasoningMistral Measured effort: Reasoning	5.0	Eligible · 100%17 published · 2 families · 1 selected capabilities	—	—	79.9108/108 peersIndependent evidence	—	53.0197/197 peersIndependent evidence	—	56.1200/200 peersIndependent evidence	—	—	19.6123/123 peersIndependent evidence	51.763/63 peersIndependent evidence	52.9212/212 peersIndependent evidence	—	52.8208/208 peersIndependent evidence	50.0178/178 peersIndependent evidence	—	83.2112/112 peersIndependent evidence	—	73.6114/114 peersIndependent evidence	37.5114/114 peersIndependent evidence	56.7198/198 peersIndependent evidence	—	39.8202/202 peersIndependent evidence	40.5201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
104	NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 · ReasoningNVIDIA Measured effort: Reasoning	4.6	Eligible · 100%20 published · 5 families · 1 selected capabilities	—	14.587/87 peersIndependent evidence	93.8108/108 peersIndependent evidence	—	40.3197/197 peersIndependent evidence	23.589/89 peersIndependent evidence	66.6200/200 peersIndependent evidence	—	12.588/88 peersIndependent evidence	23.4123/123 peersIndependent evidence	—	57.7212/212 peersIndependent evidence	—	58.2208/208 peersIndependent evidence	84.6178/178 peersIndependent evidence	—	83.1112/112 peersIndependent evidence	—	61.9114/114 peersIndependent evidence	—	41.7198/198 peersIndependent evidence	12.691/91 peersIndependent evidence	36.2202/202 peersIndependent evidence	32.4201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
105	o3-mini · HighOpenAI Measured effort: High	4.5	Eligible · 100%22 published · 5 families · 1 selected capabilities	—	18.287/87 peersIndependent evidence	—	—	41.0197/197 peersIndependent evidence	14.289/89 peersIndependent evidence	53.6200/200 peersIndependent evidence	—	25.188/88 peersIndependent evidence	21.2123/123 peersIndependent evidence	—	59.4212/212 peersIndependent evidence	—	6.2208/208 peersIndependent evidence	77.9178/178 peersIndependent evidence	—	78.9112/112 peersIndependent evidence	96.060/60 peersIndependent evidence	64.9114/114 peersIndependent evidence	—	56.5198/198 peersIndependent evidence	43.291/91 peersIndependent evidence	26.0202/202 peersIndependent evidence	23.2201/201 peersIndependent evidence	25.438/38 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
106	Mistral Small 4 · ReasoningMistral Measured effort: Reasoning	4.4	Eligible · 100%19 published · 6 families · 1 selected capabilities	0.028/28 peersIndependent evidence	22.687/87 peersIndependent evidence	—	—	51.1197/197 peersIndependent evidence	20.589/89 peersIndependent evidence	55.8200/200 peersIndependent evidence	—	1.888/88 peersIndependent evidence	30.1123/123 peersIndependent evidence	—	61.1212/212 peersIndependent evidence	—	51.4208/208 peersIndependent evidence	58.8178/178 peersIndependent evidence	—	—	—	—	32.6114/114 peersIndependent evidence	60.4198/198 peersIndependent evidence	31.691/91 peersIndependent evidence	31.6202/202 peersIndependent evidence	38.9201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
107	Qwen3-Next-80B-A3B-Thinking · ReasoningQwen Measured effort: Reasoning	4.4	Eligible · 100%16 published · 2 families · 1 selected capabilities	—	—	85.1108/108 peersIndependent evidence	—	63.7197/197 peersIndependent evidence	—	25.5200/200 peersIndependent evidence	—	—	18.5123/123 peersIndependent evidence	57.063/63 peersIndependent evidence	57.7212/212 peersIndependent evidence	—	60.2208/208 peersIndependent evidence	72.5178/178 peersIndependent evidence	—	88.8112/112 peersIndependent evidence	—	81.1114/114 peersIndependent evidence	—	49.9198/198 peersIndependent evidence	—	37.9202/202 peersIndependent evidence	34.5201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
108	Ministral 3 14B · Non-reasoningMistral Measured effort: Non-reasoning	4.3	Eligible · 100%22 published · 5 families · 1 selected capabilities	—	17.887/87 peersIndependent evidence	31.5108/108 peersIndependent evidence	—	28.9197/197 peersIndependent evidence	15.189/89 peersIndependent evidence	25.2200/200 peersIndependent evidence	—	18.688/88 peersIndependent evidence	23.5123/123 peersIndependent evidence	13.963/63 peersIndependent evidence	25.1212/212 peersIndependent evidence	—	24.1208/208 peersIndependent evidence	20.6178/178 peersIndependent evidence	—	33.9112/112 peersIndependent evidence	—	29.9114/114 peersIndependent evidence	19.8114/114 peersIndependent evidence	19.9198/198 peersIndependent evidence	6.691/91 peersIndependent evidence	29.8202/202 peersIndependent evidence	24.9201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
109	gpt-oss-20b · HighOpenAI Measured effort: High	4.3	Eligible · 100%26 published · 6 families · 1 selected capabilities	—	4.287/87 peersIndependent evidence	44.2150/150 peersIndependent evidence	0.026/26 peersIndependent evidence	23.5198/198 peersIndependent evidence	3.189/89 peersIndependent evidence	72.2200/200 peersIndependent evidence	—	25.988/88 peersIndependent evidence	27.5123/123 peersIndependent evidence	27.363/63 peersIndependent evidence	17.9213/213 peersIndependent evidence	—	55.5208/208 peersIndependent evidence	76.8178/178 peersIndependent evidence	—	87.6112/112 peersIndependent evidence	—	44.3114/114 peersIndependent evidence	—	31.5198/198 peersIndependent evidence	31.491/91 peersIndependent evidence	44.6202/202 peersIndependent evidence	33.2201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	7.850/50 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
110	Qwen3-32B · ReasoningQwen Measured effort: Reasoning	4.1	Eligible · 100%21 published · 5 families · 1 selected capabilities	—	9.287/87 peersIndependent evidence	71.3108/108 peersIndependent evidence	—	5.8197/197 peersIndependent evidence	22.489/89 peersIndependent evidence	55.7200/200 peersIndependent evidence	—	25.388/88 peersIndependent evidence	17.0123/123 peersIndependent evidence	—	39.0212/212 peersIndependent evidence	—	43.4208/208 peersIndependent evidence	31.4178/178 peersIndependent evidence	—	58.3112/112 peersIndependent evidence	74.460/60 peersIndependent evidence	63.8114/114 peersIndependent evidence	—	42.3198/198 peersIndependent evidence	23.891/91 peersIndependent evidence	27.6202/202 peersIndependent evidence	19.9201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
111	Qwen3-30B-A3B-Thinking-2507 · ReasoningQwen Measured effort: Reasoning	3.2	Eligible · 100%21 published · 5 families · 1 selected capabilities	—	12.087/87 peersIndependent evidence	56.9108/108 peersIndependent evidence	—	60.3197/197 peersIndependent evidence	17.589/89 peersIndependent evidence	55.7200/200 peersIndependent evidence	—	6.788/88 peersIndependent evidence	18.0123/123 peersIndependent evidence	—	46.1212/212 peersIndependent evidence	—	52.3208/208 peersIndependent evidence	61.6178/178 peersIndependent evidence	—	74.8112/112 peersIndependent evidence	85.560/60 peersIndependent evidence	65.9114/114 peersIndependent evidence	—	34.5198/198 peersIndependent evidence	22.191/91 peersIndependent evidence	26.1202/202 peersIndependent evidence	20.3201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
112	Ministral 3 8B · Non-reasoningMistral Measured effort: Non-reasoning	3.0	Eligible · 100%22 published · 5 families · 1 selected capabilities	—	13.587/87 peersIndependent evidence	34.7108/108 peersIndependent evidence	—	27.9197/197 peersIndependent evidence	13.289/89 peersIndependent evidence	25.2200/200 peersIndependent evidence	—	8.988/88 peersIndependent evidence	21.3123/123 peersIndependent evidence	10.563/63 peersIndependent evidence	15.9212/212 peersIndependent evidence	—	17.1208/208 peersIndependent evidence	13.8178/178 peersIndependent evidence	—	26.5112/112 peersIndependent evidence	—	18.9114/114 peersIndependent evidence	15.0114/114 peersIndependent evidence	17.0198/198 peersIndependent evidence	2.791/91 peersIndependent evidence	20.2202/202 peersIndependent evidence	21.3201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
113	Qwen3-14B · ReasoningQwen Measured effort: Reasoning	2.8	Eligible · 100%21 published · 5 families · 1 selected capabilities	—	3.987/87 peersIndependent evidence	54.8108/108 peersIndependent evidence	—	5.8197/197 peersIndependent evidence	3.089/89 peersIndependent evidence	25.2200/200 peersIndependent evidence	—	18.388/88 peersIndependent evidence	12.4123/123 peersIndependent evidence	—	29.7212/212 peersIndependent evidence	—	20.1208/208 peersIndependent evidence	43.7178/178 peersIndependent evidence	—	56.3112/112 peersIndependent evidence	75.660/60 peersIndependent evidence	53.4114/114 peersIndependent evidence	—	28.2198/198 peersIndependent evidence	14.291/91 peersIndependent evidence	31.9202/202 peersIndependent evidence	21.2201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
114	Qwen3-8B · ReasoningQwen Measured effort: Reasoning	2.8	Eligible · 100%21 published · 5 families · 1 selected capabilities	—	3.987/87 peersIndependent evidence	17.4108/108 peersIndependent evidence	—	5.8197/197 peersIndependent evidence	10.389/89 peersIndependent evidence	25.2200/200 peersIndependent evidence	—	22.588/88 peersIndependent evidence	10.0123/123 peersIndependent evidence	—	28.9212/212 peersIndependent evidence	—	11.0208/208 peersIndependent evidence	26.5178/178 peersIndependent evidence	—	41.8112/112 peersIndependent evidence	51.560/60 peersIndependent evidence	42.0114/114 peersIndependent evidence	—	20.4198/198 peersIndependent evidence	10.391/91 peersIndependent evidence	22.4202/202 peersIndependent evidence	16.4201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
115	Ministral 3 3B · Non-reasoningMistral Measured effort: Non-reasoning	2.7	Eligible · 100%22 published · 5 families · 1 selected capabilities	—	10.787/87 peersIndependent evidence	21.9108/108 peersIndependent evidence	—	20.7197/197 peersIndependent evidence	16.589/89 peersIndependent evidence	25.2200/200 peersIndependent evidence	—	2.088/88 peersIndependent evidence	16.2123/123 peersIndependent evidence	6.363/63 peersIndependent evidence	8.2212/212 peersIndependent evidence	—	34.0208/208 peersIndependent evidence	9.0178/178 peersIndependent evidence	—	17.1112/112 peersIndependent evidence	—	9.7114/114 peersIndependent evidence	5.9114/114 peersIndependent evidence	8.2198/198 peersIndependent evidence	0.091/91 peersIndependent evidence	20.1202/202 peersIndependent evidence	9.0201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
116	Magistral Small 1.2 · ReasoningMistral Measured effort: Reasoning	2.4	Eligible · 100%18 published · 3 families · 1 selected capabilities	—	—	77.8108/108 peersIndependent evidence	—	22.4197/197 peersIndependent evidence	2.889/89 peersIndependent evidence	55.9200/200 peersIndependent evidence	—	—	15.1123/123 peersIndependent evidence	28.363/63 peersIndependent evidence	36.2212/212 peersIndependent evidence	—	39.2208/208 peersIndependent evidence	53.2178/178 peersIndependent evidence	—	78.0112/112 peersIndependent evidence	—	52.5114/114 peersIndependent evidence	29.7114/114 peersIndependent evidence	19.5198/198 peersIndependent evidence	—	22.3202/202 peersIndependent evidence	21.2201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
117	GPT-4o Mini · Non-reasoningOpenAI Measured effort: Non-reasoning	2.4	Eligible · 100%12 published · 2 families · 1 selected capabilities	—	—	13.6108/108 peersIndependent evidence	—	2.657/57 peersIndependent evidence	—	—	—	—	13.1123/123 peersIndependent evidence	—	11.8212/212 peersIndependent evidence	—	16.0208/208 peersIndependent evidence	15.9178/178 peersIndependent evidence	—	16.2112/112 peersIndependent evidence	29.060/60 peersIndependent evidence	20.2114/114 peersIndependent evidence	7.4114/114 peersIndependent evidence	—	—	3.4112/112 peersIndependent evidence	17.1125/125 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
118	Mistral Small 3.2 · Non-reasoningMistral Measured effort: Non-reasoning	2.3	Eligible · 100%23 published · 5 families · 1 selected capabilities	—	4.187/87 peersIndependent evidence	28.4108/108 peersIndependent evidence	—	23.6197/197 peersIndependent evidence	9.389/89 peersIndependent evidence	25.1200/200 peersIndependent evidence	—	2.188/88 peersIndependent evidence	6.5123/123 peersIndependent evidence	21.463/63 peersIndependent evidence	18.8212/212 peersIndependent evidence	—	18.2208/208 peersIndependent evidence	26.2178/178 peersIndependent evidence	—	21.6112/112 peersIndependent evidence	45.060/60 peersIndependent evidence	25.5114/114 peersIndependent evidence	17.2114/114 peersIndependent evidence	25.0198/198 peersIndependent evidence	11.891/91 peersIndependent evidence	31.3202/202 peersIndependent evidence	26.7201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
119	NVIDIA-Nemotron-3-Nano-4B-BF16 · ReasoningNVIDIA Measured effort: Reasoning	2.2	Eligible · 100%11 published · 2 families · 1 selected capabilities	—	—	—	—	27.6197/197 peersIndependent evidence	—	25.2200/200 peersIndependent evidence	—	—	8.9123/123 peersIndependent evidence	—	20.4212/212 peersIndependent evidence	—	27.1208/208 peersIndependent evidence	70.2178/178 peersIndependent evidence	—	—	—	—	—	9.6198/198 peersIndependent evidence	—	34.9171/171 peersIndependent evidence	27.5201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
120	Llama 4 Maverick · Non-reasoningMeta Measured effort: Non-reasoning	1.8	Eligible · 100%24 published · 5 families · 1 selected capabilities	—	4.187/87 peersIndependent evidence	18.6108/108 peersIndependent evidence	—	36.3198/198 peersIndependent evidence	7.989/89 peersIndependent evidence	25.1200/200 peersIndependent evidence	—	9.088/88 peersIndependent evidence	4.6123/123 peersIndependent evidence	59.763/63 peersIndependent evidence	39.6212/212 peersIndependent evidence	—	28.1208/208 peersIndependent evidence	49.6178/178 peersIndependent evidence	—	39.3112/112 peersIndependent evidence	47.660/60 peersIndependent evidence	70.7114/114 peersIndependent evidence	42.6114/114 peersIndependent evidence	68.7198/198 peersIndependent evidence	15.791/91 peersIndependent evidence	11.9202/202 peersIndependent evidence	28.0201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
121	Gemma 4 E4B IT · ReasoningGoogle Measured effort: Reasoning	1.5	Eligible · 100%12 published · 2 families · 1 selected capabilities	—	—	—	—	33.8197/197 peersIndependent evidence	—	63.1200/200 peersIndependent evidence	—	—	11.5123/123 peersIndependent evidence	—	26.9212/212 peersIndependent evidence	—	8.9208/208 peersIndependent evidence	51.1178/178 peersIndependent evidence	—	—	—	—	22.1114/114 peersIndependent evidence	8.0198/198 peersIndependent evidence	—	20.0171/171 peersIndependent evidence	28.2201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
122	Llama-4-Scout-17B-16E-Instruct · Non-reasoningMeta Measured effort: Non-reasoning	1.5	Eligible · 100%23 published · 5 families · 1 selected capabilities	—	4.187/87 peersIndependent evidence	11.7108/108 peersIndependent evidence	—	29.7197/197 peersIndependent evidence	3.189/89 peersIndependent evidence	25.1200/200 peersIndependent evidence	—	2.188/88 peersIndependent evidence	8.3123/123 peersIndependent evidence	25.663/63 peersIndependent evidence	27.9212/212 peersIndependent evidence	—	9.2208/208 peersIndependent evidence	40.0178/178 peersIndependent evidence	—	25.6112/112 peersIndependent evidence	34.560/60 peersIndependent evidence	46.9114/114 peersIndependent evidence	24.8114/114 peersIndependent evidence	27.1198/198 peersIndependent evidence	3.991/91 peersIndependent evidence	8.7202/202 peersIndependent evidence	15.4201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
123	Gemma 3 12B IT · Non-reasoningGoogle Measured effort: Non-reasoning	1.4	Eligible · 100%22 published · 4 families · 1 selected capabilities	—	—	17.2108/108 peersIndependent evidence	—	13.4197/197 peersIndependent evidence	3.189/89 peersIndependent evidence	25.2200/200 peersIndependent evidence	—	13.088/88 peersIndependent evidence	1.9123/123 peersIndependent evidence	17.363/63 peersIndependent evidence	7.0212/212 peersIndependent evidence	—	16.2208/208 peersIndependent evidence	33.1178/178 peersIndependent evidence	—	8.1112/112 peersIndependent evidence	39.760/60 peersIndependent evidence	17.1114/114 peersIndependent evidence	4.0114/114 peersIndependent evidence	11.1198/198 peersIndependent evidence	1.391/91 peersIndependent evidence	5.9202/202 peersIndependent evidence	10.7201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
124	Gemma 4 E2B IT · ReasoningGoogle Measured effort: Reasoning	1.3	Eligible · 100%12 published · 2 families · 1 selected capabilities	—	—	—	—	20.4197/197 peersIndependent evidence	—	25.1200/200 peersIndependent evidence	—	—	7.0123/123 peersIndependent evidence	—	13.4212/212 peersIndependent evidence	—	25.0208/208 peersIndependent evidence	34.7178/178 peersIndependent evidence	—	—	—	—	13.4114/114 peersIndependent evidence	3.8198/198 peersIndependent evidence	—	20.0171/171 peersIndependent evidence	12.8201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
125	GPT-4.1-nano · Non-reasoningOpenAI Measured effort: Non-reasoning	1.3	Eligible · 100%17 published · 2 families · 1 selected capabilities	—	—	24.7108/108 peersIndependent evidence	—	24.3197/197 peersIndependent evidence	—	25.4200/200 peersIndependent evidence	—	—	5.4123/123 peersIndependent evidence	—	19.7212/212 peersIndependent evidence	—	8.2208/208 peersIndependent evidence	20.8178/178 peersIndependent evidence	—	29.9112/112 peersIndependent evidence	37.160/60 peersIndependent evidence	21.7114/114 peersIndependent evidence	6.7114/114 peersIndependent evidence	21.2198/198 peersIndependent evidence	—	10.7202/202 peersIndependent evidence	17.8201/201 peersIndependent evidence	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—

About this ranking

This profile gives agentic work all the weight. Published agent-task evaluations contribute through matched benchmark families. Harnesses and reported settings remain attached to the evidence. These results describe the evaluated configurations; they do not establish how a model will perform in every workflow.

View the Capability ranking →

Behind the ranking

273 entries with published results · 127 ranked · 0 provisional estimates.

Explore seven capabilities, the original results, and gaps in the evidence.

Coding ranking Reasoning ranking Agentic ranking Compare capabilities ↗Published benchmarks ↗Evidence coverage ↗

Cite this ranking

Include the model and effort setting, the selected profile, and the dated data downloads. A shared link uses the latest data; it does not freeze a ranking in time.

UnifyBench ranking
Data updated: 12 Sept 2026
Mode: custom; evidence: documented; configurations: best
Weights: v3;agentic:100,hard_reasoning:0,coding:0,human_pref:0,knowledge:0,multimodal:0,long_context:0
https://unifybench.ai/rankings/agentic?page=5

Catalog data · Effort measurements · Configuration scoring rules · Methodology

Catalog dated 12 Sept 2026. Effort evidence collected 12 Sept 2026. Collection dates are not evaluation dates.

What does the score mean?

The adjusted comparison score estimates performance against the same reference panel from matched published comparisons. Capability pools matched comparisons across families and checks aggregate breadth and independent corroboration. Custom profiles require support in every selected capability. Missing results never count as losses.

Family details show observed win shares against available reference peers. Those descriptive values are supporting evidence, not adjusted capability scores. Provider reports and independent results remain labeled. Matching reported settings do not establish equal inference effort, and score differences do not establish statistical significance.

Read the full methodology →