Benchmarks / Agentic

AA-Briefcase v1.1

AA-Briefcase: agentic office-work tasks over real files (xlsx/pptx/docx/pdf) judged pairwise; reported as Elo.

151models
–Open Index weight
–last updated

AA-Briefcase v1.1

Higher is better · best effort per model

Chart settings
151 of 151 selected

full table below

#ModelEffortScoreCost / taskHarness
1Sonnet 5.5max1817$20.47Stirrup
2Opus 5.5max1803$21.05Stirrup
3Muse Wonder Full (Max)max1729–Stirrup
4Fable 5.1max1672$22.71Stirrup
5Opus 5max1659$17.79Stirrup
6Muse Wondermax1659$12.45Stirrup
7Grok 4.7xhigh1643$9.30Stirrup
8Qwen 3.8 Maxmax1615$19.58Stirrup
9Muse Spark 1.3max1581$4.06Stirrup
10Qwen 3.8 Flash–1580$1.60Stirrup
11Haiku 5.5max1577$0.675Stirrup
12GPT-6 Astramax1570$9.50Stirrup
13GPT-6.1 Solmax1558$2.31Stirrup
14Grok 4.6xhigh1545$4.88Stirrup
15Fable 5max1539$22.30Stirrup
16MiMo V2.6 Pro–1516$0.380Stirrup
17GLM 5.3max1507$5.65Stirrup
18Kimi K3max1500$6.73Stirrup
19MiMo V2.6 Flash–1491$0.152Stirrup
20Gemini 4high1486$5.83Stirrup
21GPT-6 Solmax1481$2.62Stirrup
22GPT-5.6 Solmax1480$4.02Stirrup
23GLM 5.3 Flashmax1453$0.984Stirrup
24Qwen 3.8 2.4T-A95B–1435$4.16Stirrup
25Step 5 Preview–1425$2.37Stirrup
26DeepSeek V4 Flashmax1425$0.805Stirrup
27DeepSeek V4.1 Flashmax1420$0.566Stirrup
28Ling 3.1 Flash–1400$2.02Stirrup
29Qwen 3.8 27Bxhigh1397$2.83Stirrup
30Mistral Large 4–1393$2.38Stirrup
31Sonnet 5max1358$14.43Stirrup
32GPT-5.6 Lunamax1342$0.395Stirrup
33GPT-6 Lunamax1336$0.170Stirrup
34GPT-5.6 Terramax1334$2.83Stirrup
35Muse Spark 1.2xhigh1328$1.90Stirrup
36Opus 4.8max1321$8.26Stirrup
37K2 Horizon 375B A23B–1296–Stirrup
38Grok 4.5high1281$1.08Stirrup
39Muse Spark Open (Xhigh)xhigh1279$1.47Stirrup
40DeepSeek V4 Flash 0731max1259$0.423Stirrup
41DeepSeek V4 Pro 0813max1257$0.995Stirrup
42Opus 4.7max1254–Stirrup
43GLM 5.2max1229$2.82Stirrup
44Gemini 3.8 Flashhigh1203$2.70Stirrup
45K2 Horizon MoVA 36B A4B–1167–Stirrup
46GPT-5.5xhigh1137$5.15Stirrup
47Quasar 438B (Max, Based on GLM-5.2)max1123$3.76Stirrup
48Gemini 3.7 Flashhigh1114$2.15Stirrup
49Solar Pro 4–1093$3.77Stirrup
50MiniMax M3–1091$0.940Stirrup
51JT-4.1 Flash 236B A21B (Reasoning)–1089–Stirrup
52GPT-5.5 Instant–1086$1.11Stirrup
53Sonnet 4.6max1063$2.41Stirrup
54DeepSeek V4 Flash 0420high1044$0.213Stirrup
55K2 Horizon 7B–1016–Stirrup
56Apodex 1.1–1013$1.04Stirrup
57Nemotron 3.5 Super VL–999$0.703Stirrup
58Motif 3–992–Stirrup
59Ling 3.0 Flash–986–Stirrup
60Beam 1.0–976–Stirrup
61GLM 5.1–965$0.840Stirrup
62Ling 3.0 Flash Fin–961–Stirrup
63Gemini 3.6 Flashhigh954$1.56Stirrup
64Hy3–947$0.113Stirrup
65Nex-N2-Pro–941–Stirrup
66DeepSeek V4 Pro 0424max922$0.093Stirrup
67Qwen 3.7 Maxmax916$1.30Stirrup
68Qwen 3.7 Plus–915$0.240Stirrup
69Nemotron 3.5 Bolt 74B-A9B–912–Stirrup
70Inkling Small–912$0.118Stirrup
71MiMo V2.5 Pro–883$0.072Stirrup
72Gemini 3.5 Flashmedium877–Stirrup
73Nemotron 3 Ultramax876$1.33Stirrup
74Solar Mini 4–870$0.951Stirrup
75GPT-5.3 Codexxhigh868–Stirrup
76Kimi K2.7 Code–859$0.788Stirrup
77Muse Spark 1.1xhigh851$2.52Stirrup
78Inklingxhigh832–Stirrup
79Kimi K2.6–820$1.17Stirrup
80Qwen 3.6 27B–816$1.05Stirrup
81Ling 3.0 Flash–798–Stirrup
82K2 Horizon 3.7B–793–Stirrup
83Grok 4.3high762$0.298Stirrup
84MiMo V2.5–754$0.016Stirrup
85LongCat 2.0–740$0.364Stirrup
86A.X-K2–737–Stirrup
87GPT-5.4 Minixhigh719$0.374Stirrup
88MiniMax M2.7–719–Stirrup
89Sonnet 4.5–713$1.03Stirrup
90Qwen 3.6 35B-A3B–713$0.732Stirrup
91GPT-5.4 Nanoxhigh674$0.588Stirrup
92Ring-2.6-1Txhigh658$0.754Stirrup
93Gemini 3.5 Flash-Litehigh650$0.168Stirrup
94Muse Spark–646–Stirrup
95Haiku 4.5–616$0.458Stirrup
96KAT-Coder-Pro V1none601–Stirrup
97Qwen 3.5 397B-A17B–562$0.748Stirrup
98Mistral Medium 3.1none530–Stirrup
99Mistral Medium 3.5high523$0.792Stirrup
100Nemotron 3.5 Lightning–506$0.307Stirrup
101Qwen 3.5 122B-A10B–497$0.530Stirrup
102Devstral Small 2none495–Stirrup
103Devstral 2none482–Stirrup
104Muse Glimmer (High)high477$0.068Stirrup
105Ling 3.0 Tiny–475$0.0000Stirrup
106DeepSeek V3.1–458–Stirrup
107Gemini 3.1 Pro–457$1.17Stirrup
108GPT-5 Minihigh428$0.083Stirrup
109MiniCPM5-2B–423–Stirrup
110Qwen 3 Codernone404$2.16Stirrup
111Granite 4.2 30B–394$0.195Stirrup
112Mercury 2.5–378$0.110Stirrup
113Gemma 4 31B–368$0.0000Stirrup
114Command A–356$0.0000Stirrup
115Qwen 3.5 9B–341$0.243Stirrup
116Gemma 4 26B–315–Stirrup
117Mistral Small 3.1none313$0.066Stirrup
118Granite 4.2 8B–310$0.088Stirrup
119Gemini 2.5 Pro–302$0.498Stirrup
120Mercury 2high265$0.049Stirrup
121Trinity Large Thinking–255$0.153Stirrup
122Mistral Small 4–241$0.019Stirrup
123HyperNova 60B 2605 (High, Based on gpt-oss-120b)high232–Stirrup
124Mistral Large 3none228$0.051Stirrup
125North Mini Code–212$0.0000Stirrup
126Gemini 3.1 Flash-Litehigh207$0.041Stirrup
127Qwen 3 235B-A22B–169–Stirrup
128Celeris-1none151$0.146Stirrup
129o3 minihigh142–Stirrup
130Ministral 3 14Bnone131$0.052Stirrup
131DeepSeek R1–128$0.610Stirrup
132Solar Pro 3high99$0.135Stirrup
133Granite 4.2 3B–98$0.016Stirrup
134DeepSeek V3none78$0.055Stirrup
135Ministral 3 8Bnone77$0.017Stirrup
136NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)–77$0.054Stirrup
137K2 Think V2high7–Stirrup
138Qwen 3 30B-A3B–5$0.148Stirrup
139Nemotron 3 Super 120B A12B (Reasoning)–0$9.03Stirrup
140gpt-oss-120bhigh0$0.565Stirrup
141Llama 4 Mavericknone0$0.172Stirrup
142gpt-oss-20bhigh0$0.0082Stirrup
143Gemma 4 E4B (Reasoning)–0$0.014Stirrup
144Qwen 3 32B–0–Stirrup
145Mistral Small 3.2none0$0.453Stirrup
146Llama 4 Scoutnone0$0.729Stirrup
147Qwen 3 14B–0–Stirrup
148Qwen 3 8B–0–Stirrup
149Ministral 3 3Bnone0$0.022Stirrup
150NVIDIA Nemotron 3 Nano 4B–0–Stirrup
151K2 Horizon 0.9B–0–Stirrup