Qwen-AgentWorld-397B-A17B vs Inkling 2 Small

New Comparison

Scores are fetched live from the current database on every visit — nothing is cached.

Benchmark Scores Comparison

Qwen-AgentWorld-397B-A17B Inkling 2 Small

Qwen-AgentWorld-397B-A17B

Score: 15.0

Inkling 2 Small

594.1B params

Coverage matrix

2 models × 24 benchmarks — hover a benchmark name for its category

Full Coverage Matrix
Hides benchmarks where all selected models score identically.
Shading = score relative to the best score among the selected models: Low Mid High Missing (—) The totals row shows how many of the 24 benchmarks each model has scores for.
Benchmark Inkling 2 Small Thinking Machines Qwen-AgentWorld-397B-A17B Qwen Team (Alibaba Cloud)
GPQA Diamond 93 —
BioMysteryBench 70 —
AgentWorldBench Terminal — 57.73
TAU3-Bench 40.4 —
ScreenSpot Pro 83.8 —
AgentWorldBench Overall — 58.71
MuirBench 79.2 —
DeepSearch QA 90.9 —
AgentWorldBench Web — 50.98
MMMU-Pro 81.6 —
Terminal-Bench Science 0.1 5.7 —
AgentWorldBench Search — 37.82
CharXiv (RQ) 85.6 —
SWE-bench Multilingual 81.2 —
Global-MMLU-Lite 91 —
Terminal-Bench 2.1 (Terminus-2) 81.3 —
DeepSWE 72.1 —
HMMT Feb 26 95.8 —
AgentWorldBench OS — 67.89
AgentWorldBench MCP — 68.24
AgentWorldBench SWE — 68.49
SimpleVQA 68.2 —
MCP-Atlas 85 —
AgentWorldBench Android — 60.2
Covered 16/24 8/24

Scorecard

Overall 0–100 score — category strengths (50%), shared-benchmark wins (30%), coverage (20%)

Model Overall score Shared benchmark wins Coverage Strongest at Weakest at
Inkling 2 Small 66.7 — 16 of 24 (67.0%) — —
Qwen-AgentWorld-397B-A17B 33.3 — 8 of 24 (33.0%) — —

Partial coverage benchmarks

Scored by at least one but not all selected models (24) — missing scores show as —

Benchmark Category Qwen-AgentWorld-397B-A17B Inkling 2 Small
GPQA Diamond 1/2 models stem_reasoning — 93
BioMysteryBench 1/2 models stem_reasoning — 70
AgentWorldBench Terminal 1/2 models general_agent 57.73 —
TAU3-Bench 1/2 models general_agent — 40.4
ScreenSpot Pro 1/2 models general_agent — 83.8
AgentWorldBench Overall 1/2 models general_agent 58.71 —
MuirBench 1/2 models video_understanding — 79.2
DeepSearch QA 1/2 models general_agent — 90.9
AgentWorldBench Web 1/2 models general_agent 50.98 —
MMMU-Pro 1/2 models vision_language — 81.6
Terminal-Bench Science 0.1 1/2 models coding_agent — 5.7
AgentWorldBench Search 1/2 models general_agent 37.82 —
CharXiv (RQ) 1/2 models document_understanding — 85.6
SWE-bench Multilingual 1/2 models coding_agent — 81.2
Global-MMLU-Lite 1/2 models multilingual — 91
Terminal-Bench 2.1 (Terminus-2) 1/2 models coding_agent — 81.3
DeepSWE 1/2 models coding_agent — 72.1
HMMT Feb 26 1/2 models stem_reasoning — 95.8
AgentWorldBench OS 1/2 models general_agent 67.89 —
AgentWorldBench MCP 1/2 models general_agent 68.24 —
AgentWorldBench SWE 1/2 models general_agent 68.49 —
SimpleVQA 1/2 models vision_language — 68.2
MCP-Atlas 1/2 models general_agent — 85
AgentWorldBench Android 1/2 models general_agent 60.2 —

What now?

You've seen 2 models side by side — keep the loop going.

Share this comparison

The URL is the comparison — copy or share it anywhere. No tracking, no cookies.

Not the right matchup?

Swap one model out — the pre-filled form keeps the rest of the selection.

Swap Qwen-AgentWorld-397B-A17B × Swap Inkling 2 Small ×