Skip to content

Benchmark

GDPval-AA v2

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

Models and systems measured
36
Scores recorded
82
Sources
1
IQ weight
2.72

Score distribution

1,587.2
Mean
1,631
Median
253.9
Std dev
1,314
10th percentile
1,844
90th percentile
735
Min
7351,932
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.72 for chat models, measured over 34 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#Model or systemScore
1Claude Fable 5Fable/Mythos 5 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,932
2Claude Opus 4.8Fable/Mythos 5 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,890
3Claude Opus 4.8Opus 4.8 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,890
4Claude Opus 5Opus 5 System Card capability summary · adaptive thinking · max effort unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,861
5Claude Fable 5.1Anthropic Fable/Mythos 5.1 launch table· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,853
6Claude Fable 5.1System Card capability summary · adaptive thinking · max effort · default sampling · 5 trials · context ≤1M unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,853
7Claude Opus 5max· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,852
8Claude Opus 5.5Opus 5.5 System Card capability section · Artificial Analysis · max effort· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,846
9Claude Sonnet 5.5Artificial Analysis · max· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,844
10Claude Sonnet 5.5adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Artificial Analysis · Haiku 5.5 card snapshot · max· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,840
11Claude Opus 5Anthropic Fable/Mythos 5.1 launch table· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,824
12Claude Opus 5Google Model Card· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,824
13Claude Opus 5max · Artificial Analysis Stirrup · 220 tasks · shell + web browsing · human baseline 1000· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,824
14Claude Opus 5System Card capability summary · adaptive thinking · max effort · default sampling · 5 trials · context ≤1M unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,824
15Claude Opus 5.5Opus 5.5 System Card capability section · Artificial Analysis · xhigh effort· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,820
16GLM 5.3 FlashArtificial Analysis / Z.ai转载· Reported by Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,773
17GLM 5.3Artificial Analysis / Z.ai转载· Reported by Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,769
18GPT 5.5Fable/Mythos 5 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,769
19GPT 5.5Opus 4.8 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,769
20Claude Fable 5OpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,759.6
21Muse Spark 1.3max · Artificial Analysis Stirrup · 220 tasks · shell + web browsing · human baseline 1000· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,754
22Claude Opus 4.7Opus 4.8 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,753
23GPT-5.6 SolOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,747.8
24Claude Fable 5Opus 5 System Card capability summary · adaptive thinking · max effort unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,747
25Qwen3.8 MaxArtificial Analysis / Z.ai转载· Reported by Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,739
26GPT-5.6 SolOpus 5 System Card capability summary · adaptive thinking · max effort unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,736
27Claude Fable 5.1Opus 5.5 System Card capability section · Artificial Analysis · max effort· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,735
28GPT-5.6 SolArtificial Analysis / Z.ai转载· Reported by Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,730
29Claude Sonnet 5.5adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Artificial Analysis · xhigh· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,725
30Claude Fable 5Anthropic Fable/Mythos 5.1 launch table· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,723
31Claude Fable 5System Card capability summary · adaptive thinking · max effort · default sampling · 5 trials · context ≤1M unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,723
32GPT-5.6 SolAnthropic Fable/Mythos 5.1 launch table· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,711
33GPT-5.6 SolSystem Card capability summary · adaptive thinking · max effort · default sampling · 5 trials · context ≤1M unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,711
34GPT-5.6 SolGoogle Model Card· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,710
35GPT-5.6 Solmax · Artificial Analysis Stirrup · 220 tasks · shell + web browsing · human baseline 1000· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,710
36Claude Opus 5Opus 5.5 System Card capability section · Artificial Analysis · max effort· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,708
37Grok 4.7release-page snapshot· Reported by SpaceXAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,695
38Kimi K3snapshot 2026-07-23· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,686
39Kimi K3Artificial Analysis / Z.ai转载· Reported by Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,682
40Hy4 previewofficial model-card appendix· Reported by Tencent Hy· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,678
41Muse Spark 1.2Meta 官方发布图· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,631
42Claude Haiku 5.5Artificial Analysis · max· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,620
43Claude Sonnet 5Artificial Analysis · as of 2026-06-17· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,618
44Muse Spark 1.2xhigh · Artificial Analysis Stirrup · 220 tasks · shell + web browsing · human baseline 1000· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,615
45Claude Sonnet 5Gemini 3.6 Flash Model Card · July 2026· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,607
46Claude Opus 4.8OpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,600.1
47Claude Opus 4.8max · Artificial Analysis leaderboard· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,600
48Claude Opus 4.8Opus 5 System Card capability summary · adaptive thinking · max effort unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,593
49GPT-5.6 TerraOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,593
50GPT-5.6 LunaOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,591.8

Showing 1–50 of 80 results.

Not ranked

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

Model or systemScore
GPT-6 Sol (Sonnet 5.5 table label)source table label · max effort· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,487
Muse Spark (version unspecified)Meta Model API · highest effort · version unspecified · Artificial Analysis leaderboard· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,145

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.