Evidence coverage

3498 published observations across 337 measures. Capability scoring groups compatible evidence into 44 benchmark families.

A published result contributes only when a matching reference-panel result is available. Missing evidence stays unknown. Coverage describes breadth, separately from performance.

Coverage across our labs

LabModels with resultsPublished measuresLaunch sources
OpenAI34 / 4024120
Anthropic14 / 1425122
xAI4 / 8352
Google22 / 351339
Meta14 / 16462
Mistral17 / 28392
Moonshot3 / 71337
Qwen32 / 821044
DeepSeek7 / 16332
NVIDIA2 / 10471
Z.ai17 / 521057
MiniMax6 / 14692
Cohere10 / 26181
Amazon5 / 6201
Microsoft9 / 24191
Xiaomi5 / 14100
StepFun3 / 550
LG AI Research5 / 1330
Upstage5 / 940
ByteDance1 / 1610
Tencent2 / 1370
Baidu1 / 2010
OpenBMB1 / 120
Agnes AI1 / 220

Models and contributing families

Raw published counts include all source types. Contributing and independent family counts follow the selected evidence policy.

471 models match.

ModelPublished measuresContributing familiesWith independent supportExplore
Magistral Medium 1.2621Published results
MiMo-V2.5522Published results
MiMo-V2.5-Pro422Published results
MiMo-V2.6-Flash522Published results
MiMo-V2.6-Pro722Published results
Mistral Large 3722Published results
Mistral Large 42522Published results
Mistral Medium 3222Published results
Mistral Small 3.2322Published results
phi-4222Published results
Qwen3-235B-A22B-Thinking-2507222Published results
Qwen3-Coder-480B-A35B-Instruct222Published results
Qwen3.5-122B-A10B222Published results
Qwen3.5-27B222Published results
Qwen3.5-35B-A3B222Published results
Qwen3.8 Max 0902322Published results
Solar Pro 4422Published results
Step 5 Preview522Published results
C4Ai Aya Expanse 32B111Published results
chatglm-6b111Published results
chatglm2-6b111Published results
chatglm3-6b111Published results
Claude Opus 5.54811Published results
Command A 03 2025111Published results
Command A Reasoning 08 20251010Published results

Family coverage

Benchmark familyCapabilityModels contributingWith independent support
OSWorldAgentic216
Webarena VerifiedAgentic40
BrowsecompAgentic220
ToolathlonAgentic150
MCP-AtlasAgentic190
AutomationBenchAgentic210
DeepSearchQAAgentic100
GDPvalAgentic6445
Bfcl V4Agentic120
Apex AgentsAgentic100
OfficeQAAgentic110
GDP.pdfAgentic140
Terminal-Bench ScienceAgentic70
Humanity’s Last ExamHard reasoning329
GPQAHard reasoning4518
ARC-AGIHard reasoning6060
AimeHard reasoning220
FrontierMathHard reasoning100
Imo AnswerbenchHard reasoning50
SWE-bench VerifiedCoding3819
SWE-bench ProCoding204
SWE-bench MultilingualCoding90
DeepSWECoding3328
LiveCodeBenchCoding259
Terminal-BenchCoding2222
ScicodeCoding00
Human preference (Arena)Human pref119119
MMLU-ProKnowledge6852
GmmluKnowledge30
MiluKnowledge30
Simpleqa VerifiedKnowledge20
MmmuMultimodal20
Mmmu ProMultimodal240
CharXivMultimodal160
MathvistaMultimodal20
ChartographyMultimodal30
LvbenchMultimodal60
ScreenSpotMultimodal120
Ocrbench V2 Average AccuracyMultimodal120
MRCRLong context120
LongbenchLong context80
Ruler 1MLong context20
GraphwalksLong context70
Aa LcrLong context50
Model evidence matrix

Showing the same 101–125 models as the list above.

ModelOSWorldWebarena VerifiedBrowsecompToolathlonMCP-AtlasAutomationBenchDeepSearchQAGDPvalBfcl V4Apex AgentsOfficeQAGDP.pdfTerminal-Bench ScienceHumanity’s Last ExamGPQAARC-AGIAimeFrontierMathImo AnswerbenchSWE-bench VerifiedSWE-bench ProSWE-bench MultilingualDeepSWELiveCodeBenchTerminal-BenchScicodeHuman preference (Arena)MMLU-ProGmmluMiluSimpleqa VerifiedMmmuMmmu ProCharXivMathvistaChartographyLvbenchScreenSpotOcrbench V2 Average AccuracyMRCRLongbenchRuler 1MGraphwalksAa Lcr
Magistral Medium 1.2UnknownUnknown50.0
Provider reports · 2/2 reference peers
UnknownUnknownUnknownUnknown14.3
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
MiMo-V2.5UnknownUnknownUnknownUnknownUnknownUnknownUnknown57.1
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown57.9
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
MiMo-V2.5-ProUnknownUnknownUnknownUnknownUnknownUnknownUnknown100.0
Independent support · 1/1 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown84.2
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
MiMo-V2.6-FlashUnknownUnknownUnknownUnknownUnknownUnknownUnknown85.7
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown72.4
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
MiMo-V2.6-ProUnknownUnknownUnknownUnknownUnknownUnknownUnknown97.6
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown90.8
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Mistral Large 3UnknownUnknownUnknownUnknownUnknownUnknownUnknown31.0
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown47.4
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Mistral Large 4UnknownUnknownUnknownUnknownUnknownUnknownUnknown73.8
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown55.3
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Mistral Medium 3UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown0.0
Independent support · 8/8 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown68.3
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Mistral Small 3.2UnknownUnknownUnknownUnknownUnknownUnknownUnknown2.4
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown63.4
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
phi-4UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown5.3
Independent support · 76/76 reference peers
43.1
Independent support · 51/51 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Qwen3-235B-A22B-Thinking-2507UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown70.7
Independent support · 41/41 reference peers
76.5
Independent support · 51/51 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Qwen3-Coder-480B-A35B-InstructUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown41.7
Independent support · 18/18 reference peers
UnknownUnknownUnknownUnknownUnknownUnknown31.6
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Qwen3.5-122B-A10BUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown50.0
Independent support · 76/76 reference peers
90.2
Independent support · 51/51 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Qwen3.5-27BUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown42.1
Independent support · 76/76 reference peers
86.3
Independent support · 51/51 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Qwen3.5-35B-A3BUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown36.8
Independent support · 76/76 reference peers
80.4
Independent support · 51/51 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Qwen3.8 Max 0902UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown30.0
Independent support · 10/10 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown47.6
Independent support · 21/21 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Solar Pro 4UnknownUnknownUnknownUnknownUnknownUnknownUnknown69.0
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown32.9
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Step 5 PreviewUnknownUnknownUnknownUnknownUnknownUnknownUnknown81.0
Independent support · 42/42 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown71.1
Independent support · 76/76 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
C4Ai Aya Expanse 32BUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown39.0
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
chatglm-6bUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown0.0
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
chatglm2-6bUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown2.4
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
chatglm3-6bUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown4.9
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Claude Opus 5.5UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown96.6
Independent support · 59/59 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Command A 03 2025UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown61.0
Independent support · 41/41 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown
Command A Reasoning 08 2025UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown0.0
Provider reports · 1/1 reference peers
UnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknownUnknown

Family definitions and scoring rules · Compare capabilities · Compare published evidence