🏆 Tiny League

Benchmark scores for recent small open-weight AI models.

Generated 2026-10-12T17:42:24Z

Changelog

Models

Listed by release date, newest first. Click any column header to sort (Params / Benchmarks / AA-Index sort numerically, Released by date; click again to flip direction; missing values sink to the bottom). A V badge marks models that support vision / multimodal (image) input.

ModelFamilyParamsArchitectureLicenseReleasedBenchmarksAA-Index-v4.3
MiniCPM-V-4.7-35B-A3BVMiniCPM-V-4.7-35B-A3B35.2BmoeApache 2.02026-10-0733—
Agnes-3.0-FlashVAgnes-3.0-Flash33.1BdenseApache 2.02026-09-11636
Nex-N2.5-miniVNex-N2.535.1BmoeApache 2.02026-09-0816—
MiniCPM5-2BMiniCPM5-2B2.5BdenseApache 2.02026-09-063414
Ling-3.0-flash-VLVLing-3.0-flash-VL124.8BmoeMIT2026-09-041325
K2-Horizon-MoVA-36B-A4BK2-Horizon-MoVA-36B-A4B37.4BmoeApache 2.02026-09-011026
K2-Horizon-7BK2-Horizon-7B9BdenseApache 2.02026-09-01921
K2-Horizon-3.7BK2-Horizon-3.7B5.1BdenseApache 2.02026-09-01916
Qwen3.8-Flash-NextVQwen3.8180Bmoeother2026-08-242342
Spark-X2.5-4BSpark X4.1BdenseApache 2.02026-08-2421—
Ornith-1.5-9BVOrnith9.7BdenseMIT2026-08-1813—
Ornith-1.5-35B-A3BVOrnith36BmoeMIT2026-08-1815—
Ling-3.0-tinyLing7.9BmoeMIT2026-08-101812
Muse-Glimmer-30BVMuse-Glimmer-30B29.8BdenseApache 2.02026-08-092818
granite-4.2-30bgranite-4.229.3BdenseApache 2.02026-08-072015
granite-4.2-8bgranite-4.28.8BdenseApache 2.02026-08-072012
Qwen3.8-27BVQwen3.827.8BdenseApache 2.02026-08-052634
Ling-3.0-flashLing127.5BmoeMIT2026-08-022625
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16Nemotron-3.531.6Bmoeother2026-08-011514
LongCat-Flash-Lite-SparseLongCat69.1BdenseMIT2026-07-313011
KAT-Coder-V2.5-DevVKAT-Coder34.7BdenseApache 2.02026-07-237—
Nanbeige4.2-3BNanbeige4.24.2BdenseApache 2.02026-07-2121—
Laguna-S-2.1Laguna117.6Bmoeopenmdw-1.12026-07-136—
DeepSeek-V4-Flash-DSparkDeepSeek-V4165.3BmoeMIT2026-06-272235
Qwen-AgentWorld-35B-A3BVQwen-AgentWorld34.7BmoeApache 2.02026-06-221—
Laguna-XS-2.1Laguna33.4Bmoeopenmdw-1.12026-06-204—
diffusiongemma-26B-A4B-itVdiffusiongemma25.8BmoeApache 2.02026-06-091510
North-Mini-Code-1.0North30.5BmoeApache 2.02026-06-051013
LFM2.5-8B-A1BLFM2.58.5Bmoeother2026-05-28137
gemma-4-12B-itVgemma-412BdenseApache 2.02026-05-231714
Hy-MT2-30B-A3BHy-MT230.1BmoeApache 2.02026-05-119—
Nemotron-3-Nano-Omni-30B-A3B-ReasoningVNemotron-333Bmoeother2026-04-201210

Coverage heatmap

Cell shows the score for that model×benchmark; grey = absent / no numeric score. Cell colour is scaled per benchmark (light → deep blue = lower → higher); the gold cell is the best model on that benchmark. Columns are ordered by how many models report each benchmark. Click a benchmark column to rank models by it; click the Model column to sort by name.

lowerhigherbest on benchmarkVsupports vision / multimodalMoEMixture-of-Experts (MoE)
ModelTerminal-Bench
2.1
Terminal-Bench
2.0
Terminal-Bench
Hard
AA-Intelligence-…GPQA DiamondHLESWE-bench ProIFBenchSWE-bench Verifi…SciCodeAIME
2026
AIME
2025
SWE-bench Multil…LiveCodeBench
v6
LiveCodeBenchLiveCodeBench
2408-2505
AA-LCRHMMT
Feb 2026
HMMT
Feb 2025
τ³-benchBFCL
v4
BFCL
v3
BrowseCompMMLU-ProMCP-Atlasτ²-benchCharXivIMO AnswerBenchMathVisionDeepSWE
1.1
DeepSWEMMMU ProMRCRMRCR
v2
OmniDocBench 1.5GDPval-AA-V2IFEvalMulti-IFNL2Repo-BenchToolathlonVision2WebAA-Omniscience A…ClawEvalClawEval-MMCodeforces ELOGDPvalGDPval-AAJobBenchMATH500OSWorld-VerifiedWideSearchAgents' Last Exa…AndroidWorldArena-Hard-V2ArtifactsBenchBigBench Extra H…BirdBenchBrowseComp-zhCMMLUClaw-GymCoWorkBenchERQALIFEbenchLongBench-v2MMLU-ProX liteMMMLUMedXPertQA MMOCRBench 2OSWorldPinchBenchProfBenchRULERRealWorldQARecreationBenchSWE AtlasSWE-MMSkillsBenchSuperGPQAToolathlon-Verif…VideoMMEWebArena-Verifie…
Qwen3.8-27BV73··3489.230.861.779.5·····90.3············90.2·94.642.2····91.1···42.3·62.9··57.4···33.4·84.3·42.981.9·······70.765.5··········85.947.1·38.6····64.8
Qwen3.8-Flash-NextVMoE···4291.735.962.581.3····8191.9············90.6·95.758.7········48.173.564··64.4···55.7···51.284.5·······73.972.3······52.3···88.549.9·······
Qwen-AgentWorld-35B-A3BVMoE·················································································
gemma-4-12B-itV···1478.85.2····77.5··72·········77.2·69··79.7··69.1·43.40.2·········1659··········53·········83.448.7··············
diffusiongemma-26B-A4B-itVMoE···1073.211.9····69.1··69.1·········77.6·56.2··70.5··54.3·320.3·········1429··········47.6·········81.549··············
DeepSeek-V4-Flash-DSparkMoE·56.9·3588.145.152.6·79···73.3·91.6··94.8····73.286.269··88.4····78.7······47.8····3052·1395··································
Laguna-S-2.1MoE70.2·····59.4·····78.5·················40.4········49.7··································46.2······
Laguna-XS-2.1MoE·37.5····47.6·70.9···63.1····································································
LFM2.5-8B-A1BMoE···7···56.5··5042.5········49.764.8···88.1··········91.879.9···8.7······88.8································
Nanbeige4.2-3B·44.1··87.417.846.954.663.635.6···72.5··58.782.8······57.8··67.3··············52.2··74.3·············65·····················
KAT-Coder-V2.5-DevV41.0·····46.0·69.444.2··63························································93.4···········
North-Mini-Code-1.0MoE·3631.11375.711.140.257.567.638.2···70.3···································································
granite-4.2-30b29.2··1566.4·33.377.25738.8·89.241.975.8····89.26261.4··77.6·····················1225·······67.9··41.9·······66.6·····42.990.0·········
granite-4.2-8b20.6··1264.1·19.179.347.736.1·86.730.873.2····78.358.152.4··74.0·····················1189·······65.2··41.1·······61.1·····41.281.0·········
Hy-MT2-30B-A3BMoE·······50.7····························89.8············································
LongCat-Flash-Lite-Sparse·33.7·1169.5·40.6·68.2·65.7·59.3···4841.5····48.679.245.696.0·49.4····44.7···············96.8········61.984.5····53.6·················
Nemotron-3-Nano-Omni-30B-A3B-ReasoningVMoE···10······················63.6········································67.047.4··········72.2·
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16MoE24.6··1475.411.7·71.951.632.6··39.3···52··9.3··37.081.9···········832·································85.4···········
Ornith-1.5-9BV47···86.430.547.5·70.6···54.4·········56.4·54.2·············32.4···66.5·······59.5···························41.2··
Ornith-1.5-35B-A3BVMoE68.5···89.233.459.6·79···71.4·········67.6·70.2·····22·······46.248.7··72.5·······67.8·······················39.8······
Ling-3.0-tinyMoE27.7··1273.49.3·63.6·24.2······58.770.3·20.862.7······71.0·········83.2········772·······47.9·······62.3··················
Ling-3.0-flashMoE57··25·22.756.674.5··93.2·72.4··82.865.187·2873·82·65.5··83.7····90.8····87.7········1107···73.6···77·······77.3·············44.8····
Spark-X2.5-4B····67.412.344.47541.634.790.7·53.3···56.381.2·30.465.1·40.9·54.675.1·74.2········93············································
K2-Horizon-MoVA-36B-A4BMoE58.6··2680.825.2···38.9······66.3··26.8·····················18.8·······································
K2-Horizon-7B39.1··21·18.6··70.631.6······6873.3·25.8··59··························································
MiniCPM5-2B8.6··1470.28.914.466.346.426.386.586.5·69.1··5963.8·20.866.6·39.770.8·97.1·········19.686.771.8··········94.6········43.5·59.2···43.7·············40.8···
Nex-N2.5-miniVMoE73.4·····43.8···············83.4······36.1·····1446····52.9······28.5·71.2·························25.5··54.6·63.4
Ling-3.0-flash-VLVMoE···25······················81.3·84.9··79··91.3·····57.7··59.9·····································
Muse-Glimmer-30BV51.7··1883.52251.2777643.694.7·····80··23.5····75.5·78.8····74··75.8953·············65.9··························44.3····
Agnes-3.0-FlashV···3685.0··74.2·38.1······68.3························23·······································
K2-Horizon-3.7B25.1··1665.412.9··68.625.9·······70.5·17.750.9····························································
MiniCPM-V-4.7-35B-A3BVMoE·······47.7··76.5··56.6······60.1··83.5·84.775.6·74.2··69.7··························90.2········68.5·········61.2·76.4·

AA-Intelligence-Index-v4.3 vs parameters

Comparisons

Bar charts compare the top 10 models (by score) that report the same benchmark (≥2 models within the parameter range above). Hover a bar for the exact value.

Per-benchmark detail

Terminal-Bench 2.1 (16 models)

ModelScoreMetricSubtaskSourceURL
Nex-N2.5-mini73.4%Hugging Face model cardlink
Qwen3.8-27B73%TerminusHugging Face model cardlink
Laguna-S-2.170.2%Hugging Face model cardlink
Ornith-1.5-35B-A3B68.5%Claude CodeHugging Face model cardlink
K2-Horizon-MoVA-36B-A4B58.6%Hugging Face model cardlink
Ling-3.0-flash57%Hugging Face model cardlink
Muse-Glimmer-30B51.7%with terminus2Hugging Face model cardlink
Ornith-1.5-9B47%Claude CodeHugging Face model cardlink
KAT-Coder-V2.5-Dev41.0%avg of two harnessesHugging Face model cardlink
K2-Horizon-7B39.1%Hugging Face model cardlink
granite-4.2-30b29.2%Hugging Face model cardlink
Ling-3.0-tiny27.7%Hugging Face model cardlink
K2-Horizon-3.7B25.1%Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1624.6%Hugging Face model cardlink
granite-4.2-8b20.6%Hugging Face model cardlink
MiniCPM5-2B8.6%Hugging Face model cardlink

Terminal-Bench 2.0 (5 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark56.9%Hugging Face model cardlink
Nanbeige4.2-3B44.1%Hugging Face model cardlink
Laguna-XS-2.137.5%pass@1 (avg over 5 attempts)Hugging Face model cardlink
North-Mini-Code-1.036%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse33.7%w/o HIHugging Face model cardlink

Terminal-Bench Hard (1 models)

ModelScoreMetricSubtaskSourceURL
North-Mini-Code-1.031.1%Hugging Face model cardlink

AA-Intelligence-Index-v4.3 (21 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next42indexArtificial Analysislink
Agnes-3.0-Flash36pointsArtificial Analysislink
DeepSeek-V4-Flash-DSpark35indexreasoning level: Reasoning, Max Effort (highest)Artificial Analysislink
Qwen3.8-27B34indexreasoning level: xhigh (highest)Artificial Analysislink
K2-Horizon-MoVA-36B-A4B26indexArtificial Analysislink
Ling-3.0-flash25indexArtificial Analysislink
Ling-3.0-flash-VL25indexArtificial Analysislink
K2-Horizon-7B21indexArtificial Analysislink
Muse-Glimmer-30B18indexreasoning level: high (highest)Artificial Analysislink
K2-Horizon-3.7B16indexArtificial Analysislink
granite-4.2-30b15indexArtificial Analysislink
gemma-4-12B-it14indexreasoning level: Reasoning (highest)Artificial Analysislink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1614indexArtificial Analysislink
MiniCPM5-2B14indexArtificial Analysislink
North-Mini-Code-1.013indexArtificial Analysislink
granite-4.2-8b12indexArtificial Analysislink
Ling-3.0-tiny12indexArtificial Analysislink
LongCat-Flash-Lite-Sparse11indexArtificial Analysislink
diffusiongemma-26B-A4B-it10indexArtificial Analysislink
Nemotron-3-Nano-Omni-30B-A3B-Reasoning10indexArtificial Analysislink
LFM2.5-8B-A1B7indexArtificial Analysislink

GPQA Diamond (20 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next91.7%Hugging Face model cardlink
Qwen3.8-27B89.2%Hugging Face model cardlink
Ornith-1.5-35B-A3B89.2%Hugging Face model cardlink
DeepSeek-V4-Flash-DSpark88.1%Pass@1Hugging Face model cardlink
Nanbeige4.2-3B87.4%Hugging Face model cardlink
Ornith-1.5-9B86.4%Hugging Face model cardlink
Agnes-3.0-Flash85.0%Hugging Face model cardlink
Muse-Glimmer-30B83.5%high reasoningHugging Face model cardlink
K2-Horizon-MoVA-36B-A4B80.8%Hugging Face model cardlink
gemma-4-12B-it78.8%Hugging Face model cardlink
North-Mini-Code-1.075.7%OverallArtificial Analysislink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1675.4%no toolsHugging Face model cardlink
Ling-3.0-tiny73.4%Hugging Face model cardlink
diffusiongemma-26B-A4B-it73.2%Hugging Face model cardlink
MiniCPM5-2B70.2%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse69.5%avg@16, w/o HIHugging Face model cardlink
Spark-X2.5-4B67.4%Hugging Face model cardlink
granite-4.2-30b66.4%Hugging Face model cardlink
K2-Horizon-3.7B65.4%Hugging Face model cardlink
granite-4.2-8b64.1%Hugging Face model cardlink

HLE (18 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark45.1%with tools, Pass@1Hugging Face model cardlink
Qwen3.8-Flash-Next35.9%Hugging Face model cardlink
Ornith-1.5-35B-A3B33.4%with toolsHugging Face model cardlink
Qwen3.8-27B30.8%Hugging Face model cardlink
Ornith-1.5-9B30.5%with toolsHugging Face model cardlink
K2-Horizon-MoVA-36B-A4B25.2%no toolsHugging Face model cardlink
Ling-3.0-flash22.7%Hugging Face model cardlink
Muse-Glimmer-30B22%high reasoning, textHugging Face model cardlink
K2-Horizon-7B18.6%Hugging Face model cardlink
Nanbeige4.2-3B17.8%w/o searchHugging Face model cardlink
K2-Horizon-3.7B12.9%Hugging Face model cardlink
Spark-X2.5-4B12.3%Hugging Face model cardlink
diffusiongemma-26B-A4B-it11.9%with searchHugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1611.7%text-only, no toolsHugging Face model cardlink
North-Mini-Code-1.011.1%OverallArtificial Analysislink
Ling-3.0-tiny9.3%Hugging Face model cardlink
MiniCPM5-2B8.9%Hugging Face model cardlink
gemma-4-12B-it5.2%no toolsHugging Face model cardlink

SWE-bench Pro (18 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next62.5%Hugging Face model cardlink
Qwen3.8-27B61.7%Hugging Face model cardlink
Ornith-1.5-35B-A3B59.6%Hugging Face model cardlink
Laguna-S-2.159.4%Public DatasetHugging Face model cardlink
Ling-3.0-flash56.6%Hugging Face model cardlink
DeepSeek-V4-Flash-DSpark52.6%ResolvedHugging Face model cardlink
Muse-Glimmer-30B51.2%Hugging Face model cardlink
Laguna-XS-2.147.6%Public Dataset, pass@1 (avg over 2 attempts)Hugging Face model cardlink
Ornith-1.5-9B47.5%Hugging Face model cardlink
Nanbeige4.2-3B46.9%Hugging Face model cardlink
KAT-Coder-V2.5-Dev46.0%pass@1Hugging Face model cardlink
Spark-X2.5-4B44.4%Hugging Face model cardlink
Nex-N2.5-mini43.8%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse40.6%w/o HIHugging Face model cardlink
North-Mini-Code-1.040.2%OverallHugging Face model cardlink
granite-4.2-30b33.3%Hugging Face model cardlink
granite-4.2-8b19.1%Hugging Face model cardlink
MiniCPM5-2B14.4%Hugging Face model cardlink

IFBench (16 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next81.3%Hugging Face model cardlink
Qwen3.8-27B79.5%Hugging Face model cardlink
granite-4.2-8b79.3%promptHugging Face model cardlink
granite-4.2-30b77.2%promptHugging Face model cardlink
Muse-Glimmer-30B77%high reasoningHugging Face model cardlink
Spark-X2.5-4B75%Hugging Face model cardlink
Ling-3.0-flash74.5%Hugging Face model cardlink
Agnes-3.0-Flash74.2%Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1671.9%looseHugging Face model cardlink
MiniCPM5-2B66.3%Hugging Face model cardlink
Ling-3.0-tiny63.6%Hugging Face model cardlink
North-Mini-Code-1.057.5%OverallArtificial Analysislink
LFM2.5-8B-A1B56.5%Hugging Face model cardlink
Nanbeige4.2-3B54.6%Hugging Face model cardlink
Hy-MT2-30B-A3B50.7%OverallOfficial tech report / arXiv paperlink
MiniCPM-V-4.7-35B-A3B47.7%instructHugging Face model cardlink

SWE-bench Verified (16 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark79%ResolvedHugging Face model cardlink
Ornith-1.5-35B-A3B79%Hugging Face model cardlink
Muse-Glimmer-30B76%Hugging Face model cardlink
Laguna-XS-2.170.9%pass@1 (avg over 4 attempts)Hugging Face model cardlink
Ornith-1.5-9B70.6%Hugging Face model cardlink
K2-Horizon-7B70.6%Hugging Face model cardlink
KAT-Coder-V2.5-Dev69.4%pass@1Hugging Face model cardlink
K2-Horizon-3.7B68.6%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse68.2%w/o HIHugging Face model cardlink
North-Mini-Code-1.067.6%OverallHugging Face model cardlink
Nanbeige4.2-3B63.6%Hugging Face model cardlink
granite-4.2-30b57%Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1651.6%OverallHugging Face model cardlink
granite-4.2-8b47.7%Hugging Face model cardlink
MiniCPM5-2B46.4%Hugging Face model cardlink
Spark-X2.5-4B41.6%Hugging Face model cardlink

SciCode (14 models)

ModelScoreMetricSubtaskSourceURL
KAT-Coder-V2.5-Dev44.2%pass@1Hugging Face model cardlink
Muse-Glimmer-30B43.6%Hugging Face model cardlink
K2-Horizon-MoVA-36B-A4B38.9%OverallHugging Face model cardlink
granite-4.2-30b38.8%Hugging Face model cardlink
North-Mini-Code-1.038.2%OverallHugging Face model cardlink
Agnes-3.0-Flash38.1%Hugging Face model cardlink
granite-4.2-8b36.1%Hugging Face model cardlink
Nanbeige4.2-3B35.6%Hugging Face model cardlink
Spark-X2.5-4B34.7%Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1632.6%OverallHugging Face model cardlink
K2-Horizon-7B31.6%OverallHugging Face model cardlink
MiniCPM5-2B26.3%Hugging Face model cardlink
K2-Horizon-3.7B25.9%OverallHugging Face model cardlink
Ling-3.0-tiny24.2%Hugging Face model cardlink

AIME 2026 (9 models)

ModelScoreMetricSubtaskSourceURL
Muse-Glimmer-30B94.7%high reasoningHugging Face model cardlink
Ling-3.0-flash93.2%Hugging Face model cardlink
Spark-X2.5-4B90.7%Hugging Face model cardlink
MiniCPM5-2B86.5%Hugging Face model cardlink
gemma-4-12B-it77.5%no toolsHugging Face model cardlink
MiniCPM-V-4.7-35B-A3B76.5%avg@16, instructHugging Face model cardlink
diffusiongemma-26B-A4B-it69.1%no toolsHugging Face model cardlink
LongCat-Flash-Lite-Sparse65.7%avg@32, w/o HIHugging Face model cardlink
LFM2.5-8B-A1B50%Hugging Face model cardlink

AIME 2025 (4 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b89.2%Hugging Face model cardlink
granite-4.2-8b86.7%Hugging Face model cardlink
MiniCPM5-2B86.5%Hugging Face model cardlink
LFM2.5-8B-A1B42.5%Hugging Face model cardlink

SWE-bench Multilingual (13 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next81%Hugging Face model cardlink
Laguna-S-2.178.5%Hugging Face model cardlink
DeepSeek-V4-Flash-DSpark73.3%ResolvedHugging Face model cardlink
Ling-3.0-flash72.4%Hugging Face model cardlink
Ornith-1.5-35B-A3B71.4%Hugging Face model cardlink
Laguna-XS-2.163.1%pass@1 (avg over 4 attempts)Hugging Face model cardlink
KAT-Coder-V2.5-Dev63%pass@1Hugging Face model cardlink
LongCat-Flash-Lite-Sparse59.3%w/o HIHugging Face model cardlink
Ornith-1.5-9B54.4%Hugging Face model cardlink
Spark-X2.5-4B53.3%Hugging Face model cardlink
granite-4.2-30b41.9%Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1639.3%OverallHugging Face model cardlink
granite-4.2-8b30.8%Hugging Face model cardlink

LiveCodeBench v6 (10 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next91.9%Hugging Face model cardlink
Qwen3.8-27B90.3%Hugging Face model cardlink
granite-4.2-30b75.8%Hugging Face model cardlink
granite-4.2-8b73.2%Hugging Face model cardlink
Nanbeige4.2-3B72.5%Hugging Face model cardlink
gemma-4-12B-it72%Hugging Face model cardlink
North-Mini-Code-1.070.3%Hugging Face model cardlink
diffusiongemma-26B-A4B-it69.1%Hugging Face model cardlink
MiniCPM5-2B69.1%Hugging Face model cardlink
MiniCPM-V-4.7-35B-A3B56.6%instructHugging Face model cardlink

LiveCodeBench (1 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark91.6%Pass@1Hugging Face model cardlink

LiveCodeBench 2408-2505 (1 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash82.8%Hugging Face model cardlink

AA-LCR (11 models)

ModelScoreMetricSubtaskSourceURL
Muse-Glimmer-30B80%Hugging Face model cardlink
Agnes-3.0-Flash68.3%Hugging Face model cardlink
K2-Horizon-7B68%OverallHugging Face model cardlink
K2-Horizon-MoVA-36B-A4B66.3%OverallHugging Face model cardlink
Ling-3.0-flash65.1%Hugging Face model cardlink
MiniCPM5-2B59%Hugging Face model cardlink
Nanbeige4.2-3B58.7%Hugging Face model cardlink
Ling-3.0-tiny58.7%Hugging Face model cardlink
Spark-X2.5-4B56.3%Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1652idxOverallHugging Face model cardlink
LongCat-Flash-Lite-Sparse48%w/o HIHugging Face model cardlink

HMMT Feb 2026 (9 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark94.8%Pass@1Hugging Face model cardlink
Ling-3.0-flash87%Hugging Face model cardlink
Nanbeige4.2-3B82.8%Hugging Face model cardlink
Spark-X2.5-4B81.2%Hugging Face model cardlink
K2-Horizon-7B73.3%Hugging Face model cardlink
K2-Horizon-3.7B70.5%Hugging Face model cardlink
Ling-3.0-tiny70.3%Hugging Face model cardlink
MiniCPM5-2B63.8%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse41.5%avg@32, w/ HIHugging Face model cardlink

HMMT Feb 2025 (2 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b89.2%Hugging Face model cardlink
granite-4.2-8b78.3%Hugging Face model cardlink

τ³-bench (11 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b62%avgHugging Face model cardlink
granite-4.2-8b58.1%avgHugging Face model cardlink
Spark-X2.5-4B30.4%Hugging Face model cardlink
Ling-3.0-flash28%BankingHugging Face model cardlink
K2-Horizon-MoVA-36B-A4B26.8%BankingHugging Face model cardlink
K2-Horizon-7B25.8%BankingHugging Face model cardlink
Muse-Glimmer-30B23.5%BankingHugging Face model cardlink
Ling-3.0-tiny20.8%BankingHugging Face model cardlink
MiniCPM5-2B20.8%BankingHugging Face model cardlink
K2-Horizon-3.7B17.7%BankingHugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF169.3%BankingHugging Face model cardlink

BFCL v4 (9 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash73%Hugging Face model cardlink
MiniCPM5-2B66.6%Hugging Face model cardlink
Spark-X2.5-4B65.1%Hugging Face model cardlink
Ling-3.0-tiny62.7%Hugging Face model cardlink
granite-4.2-30b61.4%Hugging Face model cardlink
MiniCPM-V-4.7-35B-A3B60.1%instructHugging Face model cardlink
granite-4.2-8b52.4%Hugging Face model cardlink
K2-Horizon-3.7B50.9%Hugging Face model cardlink
LFM2.5-8B-A1B49.7%Hugging Face model cardlink

BFCL v3 (1 models)

ModelScoreMetricSubtaskSourceURL
LFM2.5-8B-A1B64.8%Hugging Face model cardlink

BrowseComp (10 models)

ModelScoreMetricSubtaskSourceURL
Nex-N2.5-mini83.4%Hugging Face model cardlink
Ling-3.0-flash82%multi-agentHugging Face model cardlink
DeepSeek-V4-Flash-DSpark73.2%Pass@1Hugging Face model cardlink
Ornith-1.5-35B-A3B67.6%Hugging Face model cardlink
K2-Horizon-7B59%Hugging Face model cardlink
Ornith-1.5-9B56.4%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse48.6%pass@1, w/o HIHugging Face model cardlink
Spark-X2.5-4B40.9%Hugging Face model cardlink
MiniCPM5-2B39.7%Top100Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1637.0%OverallHugging Face model cardlink

MMLU-Pro (9 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark86.2%EMHugging Face model cardlink
MiniCPM-V-4.7-35B-A3B83.5%instructHugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1681.9%OverallHugging Face model cardlink
LongCat-Flash-Lite-Sparse79.2%w/o HIHugging Face model cardlink
diffusiongemma-26B-A4B-it77.6%Hugging Face model cardlink
granite-4.2-30b77.6%Hugging Face model cardlink
gemma-4-12B-it77.2%Hugging Face model cardlink
granite-4.2-8b74.0%Hugging Face model cardlink
MiniCPM5-2B70.8%Hugging Face model cardlink

MCP-Atlas (8 models)

ModelScoreMetricSubtaskSourceURL
Muse-Glimmer-30B75.5%PublicHugging Face model cardlink
Ornith-1.5-35B-A3B70.2%Hugging Face model cardlink
DeepSeek-V4-Flash-DSpark69%Pass@1Hugging Face model cardlink
Ling-3.0-flash65.5%Hugging Face model cardlink
Nanbeige4.2-3B57.8%Hugging Face model cardlink
Spark-X2.5-4B54.6%Hugging Face model cardlink
Ornith-1.5-9B54.2%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse45.6%w/o HIHugging Face model cardlink

τ²-bench (7 models)

ModelScoreMetricSubtaskSourceURL
MiniCPM5-2B97.1%TelecomHugging Face model cardlink
LongCat-Flash-Lite-Sparse96.0%Telecom, avg@4, w/ HIHugging Face model cardlink
LFM2.5-8B-A1B88.1%TelecomHugging Face model cardlink
MiniCPM-V-4.7-35B-A3B84.7%instructHugging Face model cardlink
Spark-X2.5-4B75.1%Hugging Face model cardlink
gemma-4-12B-it69%average over 3Hugging Face model cardlink
diffusiongemma-26B-A4B-it56.2%average over 3Hugging Face model cardlink

CharXiv (6 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next90.6%RQ, with CIHugging Face model cardlink
Qwen3.8-27B90.2%RQ with CIHugging Face model cardlink
Ling-3.0-flash-VL81.3%RQHugging Face model cardlink
Muse-Glimmer-30B78.8%ReasoningHugging Face model cardlink
MiniCPM-V-4.7-35B-A3B75.6%Reasoning, 4x compression, instructHugging Face model cardlink
Nemotron-3-Nano-Omni-30B-A3B-Reasoning63.6%Chart ReasoningHugging Face model cardlink

IMO AnswerBench (6 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark88.4%Pass@1Hugging Face model cardlink
Ling-3.0-flash83.7%Hugging Face model cardlink
Spark-X2.5-4B74.2%Hugging Face model cardlink
Ling-3.0-tiny71.0%Hugging Face model cardlink
Nanbeige4.2-3B67.3%Hugging Face model cardlink
LongCat-Flash-Lite-Sparse49.4%avg@4, w/o HIHugging Face model cardlink

MathVision (6 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next95.7%with CIHugging Face model cardlink
Qwen3.8-27B94.6%with CIHugging Face model cardlink
Ling-3.0-flash-VL84.9%Hugging Face model cardlink
gemma-4-12B-it79.7%Hugging Face model cardlink
MiniCPM-V-4.7-35B-A3B74.2%4x visual token compression, instructHugging Face model cardlink
diffusiongemma-26B-A4B-it70.5%Hugging Face model cardlink

DeepSWE 1.1 (3 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next58.7%Hugging Face model cardlink
Qwen3.8-27B42.2%Hugging Face model cardlink
Nex-N2.5-mini36.1%Hugging Face model cardlink

DeepSWE (2 models)

ModelScoreMetricSubtaskSourceURL
Laguna-S-2.140.4%Hugging Face model cardlink
Ornith-1.5-35B-A3B22%Hugging Face model cardlink

MMMU Pro (5 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash-VL79%Hugging Face model cardlink
Muse-Glimmer-30B74%Hugging Face model cardlink
MiniCPM-V-4.7-35B-A3B69.7%4x visual token compression, instructHugging Face model cardlink
gemma-4-12B-it69.1%Hugging Face model cardlink
diffusiongemma-26B-A4B-it54.3%Hugging Face model cardlink

MRCR (3 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash90.8%128KHugging Face model cardlink
DeepSeek-V4-Flash-DSpark78.7%1MHugging Face model cardlink
LongCat-Flash-Lite-Sparse44.7%8-needle, w/o HIHugging Face model cardlink

MRCR v2 (2 models)

ModelScoreMetricSubtaskSourceURL
gemma-4-12B-it43.4%8 needle 128k (average)Hugging Face model cardlink
diffusiongemma-26B-A4B-it32%8 needle 128k (average)Hugging Face model cardlink

OmniDocBench 1.5 (5 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash-VL91.3%Hugging Face model cardlink
Qwen3.8-27B91.1%Hugging Face model cardlink
Muse-Glimmer-30B75.8%Hugging Face model cardlink
diffusiongemma-26B-A4B-it0.3avg edit distanceHugging Face model cardlink
gemma-4-12B-it0.2avg edit distanceHugging Face model cardlink

GDPval-AA-V2 (4 models)

ModelScoreMetricSubtaskSourceURL
Nex-N2.5-mini1446pointsOverallHugging Face model cardlink
Muse-Glimmer-30B953pointsOverallHugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16832pointsOverallHugging Face model cardlink
MiniCPM5-2B19.6%Hugging Face model cardlink

IFEval (4 models)

ModelScoreMetricSubtaskSourceURL
Spark-X2.5-4B93%Hugging Face model cardlink
LFM2.5-8B-A1B91.8%Hugging Face model cardlink
Hy-MT2-30B-A3B89.8%OverallOfficial tech report / arXiv paperlink
MiniCPM5-2B86.7%Hugging Face model cardlink

Multi-IF (4 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash87.7%Hugging Face model cardlink
Ling-3.0-tiny83.2%Hugging Face model cardlink
LFM2.5-8B-A1B79.9%Hugging Face model cardlink
MiniCPM5-2B71.8%Hugging Face model cardlink

NL2Repo-Bench (4 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next48.1%Hugging Face model cardlink
Ornith-1.5-35B-A3B46.2%Hugging Face model cardlink
Qwen3.8-27B42.3%Hugging Face model cardlink
Ornith-1.5-9B32.4%Hugging Face model cardlink

Toolathlon (4 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next73.5%Verified, Pass@1Hugging Face model cardlink
Laguna-S-2.149.7%VerifiedHugging Face model cardlink
Ornith-1.5-35B-A3B48.7%VerifiedHugging Face model cardlink
DeepSeek-V4-Flash-DSpark47.8%Pass@1Hugging Face model cardlink

Vision2Web (4 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next64%Hugging Face model cardlink
Qwen3.8-27B62.9%Hugging Face model cardlink
Ling-3.0-flash-VL57.7%Hugging Face model cardlink
Nex-N2.5-mini52.9%Hugging Face model cardlink

AA-Omniscience Accuracy (3 models)

ModelScoreMetricSubtaskSourceURL
Agnes-3.0-Flash23%Hugging Face model cardlink
K2-Horizon-MoVA-36B-A4B18.8%Hugging Face model cardlink
LFM2.5-8B-A1B8.7%Hugging Face model cardlink

ClawEval (3 models)

ModelScoreMetricSubtaskSourceURL
Ornith-1.5-35B-A3B72.5%Hugging Face model cardlink
Ornith-1.5-9B66.5%Hugging Face model cardlink
Nanbeige4.2-3B52.2%pass^3Hugging Face model cardlink

ClawEval-MM (3 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next64.4%Pass@3Hugging Face model cardlink
Ling-3.0-flash-VL59.9%Hugging Face model cardlink
Qwen3.8-27B57.4%pass@3Hugging Face model cardlink

Codeforces ELO (3 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark3052ratingHugging Face model cardlink
gemma-4-12B-it1659ratingHugging Face model cardlink
diffusiongemma-26B-A4B-it1429ratingHugging Face model cardlink

GDPval (3 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b1225scoreHugging Face model cardlink
granite-4.2-8b1189pointsHugging Face model cardlink
Nanbeige4.2-3B74.3%Hugging Face model cardlink

GDPval-AA (3 models)

ModelScoreMetricSubtaskSourceURL
DeepSeek-V4-Flash-DSpark1395EloHugging Face model cardlink
Ling-3.0-flash1107pointsv2Hugging Face model cardlink
Ling-3.0-tiny772pointsv2Hugging Face model cardlink

JobBench (3 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next55.7%Hugging Face model cardlink
Qwen3.8-27B33.4%Hugging Face model cardlink
Nex-N2.5-mini28.5%Hugging Face model cardlink

MATH500 (3 models)

ModelScoreMetricSubtaskSourceURL
LongCat-Flash-Lite-Sparse96.8%w/ HIHugging Face model cardlink
MiniCPM5-2B94.6%Hugging Face model cardlink
LFM2.5-8B-A1B88.8%Hugging Face model cardlink

OSWorld-Verified (3 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-27B84.3%Hugging Face model cardlink
Nex-N2.5-mini71.2%NexCUA harnessHugging Face model cardlink
Muse-Glimmer-30B65.9%Hugging Face model cardlink

WideSearch (3 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash73.6%Hugging Face model cardlink
Ornith-1.5-35B-A3B67.8%Hugging Face model cardlink
Ornith-1.5-9B59.5%Hugging Face model cardlink

Agents' Last Exam (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next51.2scoreScoreHugging Face model cardlink
Qwen3.8-27B42.9%scoreHugging Face model cardlink

AndroidWorld (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next84.5%Hugging Face model cardlink
Qwen3.8-27B81.9%Hugging Face model cardlink

Arena-Hard-V2 (2 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b67.9%Hugging Face model cardlink
granite-4.2-8b65.2%Hugging Face model cardlink

ArtifactsBench (2 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash77%Hugging Face model cardlink
Ling-3.0-tiny47.9%Hugging Face model cardlink

BigBench Extra Hard (2 models)

ModelScoreMetricSubtaskSourceURL
gemma-4-12B-it53%Hugging Face model cardlink
diffusiongemma-26B-A4B-it47.6%Hugging Face model cardlink

BirdBench (2 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b41.9%Hugging Face model cardlink
granite-4.2-8b41.1%Hugging Face model cardlink

BrowseComp-zh (2 models)

ModelScoreMetricSubtaskSourceURL
LongCat-Flash-Lite-Sparse61.9%pass@1, w/o HIHugging Face model cardlink
MiniCPM5-2B43.5%Hugging Face model cardlink

CMMLU (2 models)

ModelScoreMetricSubtaskSourceURL
MiniCPM-V-4.7-35B-A3B90.2%instructHugging Face model cardlink
LongCat-Flash-Lite-Sparse84.5%w/ HIHugging Face model cardlink

Claw-Gym (2 models)

ModelScoreMetricSubtaskSourceURL
Nanbeige4.2-3B65%Hugging Face model cardlink
MiniCPM5-2B59.2%Hugging Face model cardlink

CoWorkBench (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next73.9%Hugging Face model cardlink
Qwen3.8-27B70.7%Hugging Face model cardlink

ERQA (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next72.3%Hugging Face model cardlink
Qwen3.8-27B65.5%Hugging Face model cardlink

LIFEbench (2 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash77.3%Hugging Face model cardlink
Ling-3.0-tiny62.3%Hugging Face model cardlink

LongBench-v2 (2 models)

ModelScoreMetricSubtaskSourceURL
LongCat-Flash-Lite-Sparse53.6%w/ HIHugging Face model cardlink
MiniCPM5-2B43.7%Hugging Face model cardlink

MMLU-ProX lite (2 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b66.6%Hugging Face model cardlink
granite-4.2-8b61.1%Hugging Face model cardlink

MMMLU (2 models)

ModelScoreMetricSubtaskSourceURL
gemma-4-12B-it83.4%Hugging Face model cardlink
diffusiongemma-26B-A4B-it81.5%Hugging Face model cardlink

MedXPertQA MM (2 models)

ModelScoreMetricSubtaskSourceURL
diffusiongemma-26B-A4B-it49%Hugging Face model cardlink
gemma-4-12B-it48.7%Hugging Face model cardlink

OCRBench 2 (2 models)

ModelScoreMetricSubtaskSourceURL
MiniCPM-V-4.7-35B-A3B68.5%4x visual token compression, instructHugging Face model cardlink
Nemotron-3-Nano-Omni-30B-A3B-Reasoning67.0%ENHugging Face model cardlink

OSWorld (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next52.3%2.0, PartialHugging Face model cardlink
Nemotron-3-Nano-Omni-30B-A3B-Reasoning47.4%Computer UseHugging Face model cardlink

PinchBench (2 models)

ModelScoreMetricSubtaskSourceURL
KAT-Coder-V2.5-Dev93.4%pass@1Hugging Face model cardlink
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF1685.4%OverallHugging Face model cardlink

ProfBench (2 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b42.9%Hugging Face model cardlink
granite-4.2-8b41.2%Hugging Face model cardlink

RULER (2 models)

ModelScoreMetricSubtaskSourceURL
granite-4.2-30b90.0%64KHugging Face model cardlink
granite-4.2-8b81.0%64KHugging Face model cardlink

RealWorldQA (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next88.5%Hugging Face model cardlink
Qwen3.8-27B85.9%Hugging Face model cardlink

RecreationBench (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-Flash-Next49.9%Hugging Face model cardlink
Qwen3.8-27B47.1%Hugging Face model cardlink

SWE Atlas (2 models)

ModelScoreMetricSubtaskSourceURL
Laguna-S-2.146.2%Codebase QnAHugging Face model cardlink
Ornith-1.5-35B-A3B39.8%QnAHugging Face model cardlink

SWE-MM (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-27B38.6%Hugging Face model cardlink
Nex-N2.5-mini25.5%Hugging Face model cardlink

SkillsBench (2 models)

ModelScoreMetricSubtaskSourceURL
Ling-3.0-flash44.8%Hugging Face model cardlink
Muse-Glimmer-30B44.3%with skillsHugging Face model cardlink

SuperGPQA (2 models)

ModelScoreMetricSubtaskSourceURL
MiniCPM-V-4.7-35B-A3B61.2%instructHugging Face model cardlink
MiniCPM5-2B40.8%Hugging Face model cardlink

Toolathlon-Verified (2 models)

ModelScoreMetricSubtaskSourceURL
Nex-N2.5-mini54.6%Hugging Face model cardlink
Ornith-1.5-9B41.2%Hugging Face model cardlink

VideoMME (2 models)

ModelScoreMetricSubtaskSourceURL
MiniCPM-V-4.7-35B-A3B76.4%w/o subtitles, 4x compression, instructHugging Face model cardlink
Nemotron-3-Nano-Omni-30B-A3B-Reasoning72.2%Video Q/AHugging Face model cardlink

WebArena-Verified (2 models)

ModelScoreMetricSubtaskSourceURL
Qwen3.8-27B64.8%Hugging Face model cardlink
Nex-N2.5-mini63.4%NexCUA harnessHugging Face model cardlink