Skip to content

Benchmark

AppWorld Test-Normal Level 1: 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

67%
Mean
75.4%
Median
28.3%
Std dev
26.3%
10th percentile
98.2%
90th percentile
5.3%
Min
5.3%100%
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 docs82.5%
claude-opus-4-7LARA MAS (23715209-d5c9ade3)· AppWorld Leaderboard data docs98.2%
DeepSeekCoderFullCodeRefl (0465049f-bff9fca2)· AppWorld Leaderboard data docs26.3%
DeepSeekCoderPlanExec (871bf9fd-4de66f57)· AppWorld Leaderboard data docs5.3%
DeepSeekCoderReAct (9ae299b8-25c255f5)· AppWorld Leaderboard data docs21.1%
GLM-5.3AtlasHarness (65df675b-ca5e4771)· AppWorld Leaderboard data docs98.2%
GLM-5.3-FlashIBM CUGA (a2db3f1e-2c28b1b2)· AppWorld Leaderboard data docs100%
GPT-4 TurboFullCodeRefl (571aa685-a8671c57)· AppWorld Leaderboard data docs43.9%
GPT-4 TurboIPFunCall (08d2ff3e-a0589430)· AppWorld Leaderboard data docs43.9%
GPT-4 TurboPlanExec (c722a232-74b9384b)· AppWorld Leaderboard data docs64.9%
GPT-4 TurboReAct (1f0055c0-123849bb)· AppWorld Leaderboard data docs50.9%
GPT-4.1IBM CUGA (8a02d49a-0df800d6)· AppWorld Leaderboard data docs91.2%
gpt-4.1-miniLARA MAS (695a4b55-1b68dde4)· AppWorld Leaderboard data docs91.2%
GPT-4oFullCodeRefl (d035381e-2d30c0e3)· AppWorld Leaderboard data docs61.4%
GPT-4oIPFunCall (3a889687-923696fe)· AppWorld Leaderboard data docs49.1%
GPT-4oPlanExec (ac5b742b-aa0243da)· AppWorld Leaderboard data docs75.4%
GPT-4oReAct (3414a75f-3df6a61f)· AppWorld Leaderboard data docs73.7%
GPT-4oReAct + 2 SetBSR Demos (1c12639a-57a0ac34)· AppWorld Leaderboard data docs87.7%
gpt-5.6-lunakecaipan capybara (1095b768-6513e0c6)· AppWorld Leaderboard data docs94.7%
gptoss-20bRooAGI_agentbuilder_memtune_rl (5acab4bb-badb978f)· AppWorld Leaderboard data docs91.2%
LLaMA3-70BFullCodeRefl (baf95530-cee120fc)· AppWorld Leaderboard data docs50.9%
LLaMA3-70BPlanExec (77f25849-862250c4)· AppWorld Leaderboard data docs19.3%
LLaMA3-70BReAct (4077db41-07bc5204)· AppWorld Leaderboard data docs42.1%
Qwen2.5-32BLOOP (e5e08b7e-d8e6474f)· AppWorld Leaderboard data docs87.7%
Qwen3-14BAlibaba Cloud ApsaraLab AgentRL (cabab065-0caee3c3)· AppWorld Leaderboard data docs100%
Qwen3.5-27BSynthAPI (3bd98253-d4fa6c8e)· AppWorld Leaderboard data docs91.2%

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