Skip to content

Benchmark

AppWorld Test-Normal: Task Goal Completion

Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages. Higher is better.

Models and systems measured
26
Scores recorded
26
Sources
1
IQ weight
No weight published

Score distribution

51.2%
Mean
48.8%
Median
30.7%
Std dev
13.1%
10th percentile
91.1%
90th percentile
1.8%
Min
1.8%94%
The band covers the middle 80% of scores. The line marks the median.

Results

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or systemScore
Claude Opus 4.6Button Agent (f072cd07-6ae0fbdf)· AppWorld Leaderboard data docs91.1%
claude-opus-4-7LARA MAS (23715209-d5c9ade3)· AppWorld Leaderboard data docs88.7%
DeepSeekCoderFullCodeRefl (0465049f-bff9fca2)· AppWorld Leaderboard data docs13.1%
DeepSeekCoderPlanExec (871bf9fd-4de66f57)· AppWorld Leaderboard data docs1.8%
DeepSeekCoderReAct (9ae299b8-25c255f5)· AppWorld Leaderboard data docs7.1%
GLM-5.3AtlasHarness (65df675b-ca5e4771)· AppWorld Leaderboard data docs94%
GLM-5.3-FlashIBM CUGA (a2db3f1e-2c28b1b2)· AppWorld Leaderboard data docs93.5%
GPT-4 TurboFullCodeRefl (571aa685-a8671c57)· AppWorld Leaderboard data docs25.6%
GPT-4 TurboIPFunCall (08d2ff3e-a0589430)· AppWorld Leaderboard data docs30.4%
GPT-4 TurboPlanExec (c722a232-74b9384b)· AppWorld Leaderboard data docs32.7%
GPT-4 TurboReAct (1f0055c0-123849bb)· AppWorld Leaderboard data docs26.8%
GPT-4.1IBM CUGA (8a02d49a-0df800d6)· AppWorld Leaderboard data docs73.2%
gpt-4.1-miniLARA MAS (695a4b55-1b68dde4)· AppWorld Leaderboard data docs61.9%
GPT-4oFullCodeRefl (d035381e-2d30c0e3)· AppWorld Leaderboard data docs33.9%
GPT-4oIPFunCall (3a889687-923696fe)· AppWorld Leaderboard data docs32.1%
GPT-4oPlanExec (ac5b742b-aa0243da)· AppWorld Leaderboard data docs44.6%
GPT-4oReAct (3414a75f-3df6a61f)· AppWorld Leaderboard data docs48.8%
GPT-4oReAct + 2 SetBSR Demos (1c12639a-57a0ac34)· AppWorld Leaderboard data docs68.5%
gpt-5.6-lunakecaipan capybara (1095b768-6513e0c6)· AppWorld Leaderboard data docs85.1%
gptoss-20bRooAGI_agentbuilder_memtune_rl (5acab4bb-badb978f)· AppWorld Leaderboard data docs76.2%
LLaMA3-70BFullCodeRefl (baf95530-cee120fc)· AppWorld Leaderboard data docs24.4%
LLaMA3-70BPlanExec (77f25849-862250c4)· AppWorld Leaderboard data docs8.9%
LLaMA3-70BReAct (4077db41-07bc5204)· AppWorld Leaderboard data docs20.8%
Qwen2.5-32BLOOP (e5e08b7e-d8e6474f)· AppWorld Leaderboard data docs72.6%
Qwen3-14BAlibaba Cloud ApsaraLab AgentRL (cabab065-0caee3c3)· AppWorld Leaderboard data docs86.9%
Qwen3.5-27BSynthAPI (3bd98253-d4fa6c8e)· AppWorld Leaderboard data docs88.7%

Data credits: AppWorld leaderboard (Stony Brook NLP). Aggregate scores from the official leaderboard; no results licence asserted. Protected bundles and experiment outputs are not used