AI EVALUATION, IN CONTEXT.BY EQUALITY MACHINE
Benchmark Radar.Benchmarks. Model results. Original sources.
/

Artificial Analysis

APEX-Agents-AA

Long-horizon agentic tasks

Evaluation context and evidence

Source
Artificial Analysis
Publisher
Not recorded
Benchmark release
Not recorded
Collected
2026-08-25
Modality
Not recorded

These are source-reported observations, not independent reruns by this site. Check the task version, evaluation protocol, unit and reporting date before comparing models. Results from different setups are not a universal ranking.

Original sources

Reported results 29

Shown in source observation order, not ranked. Dates retain the precision provided by the source.

ModelValueUnitProtocolReported date and basisSource
Gemini 3.5 Flash (high)0.470501474926254Not recordedNot recorded2026-05-19 · model_announcementArtificial Analysis
Kimi K3 (max)0.412979351032448Not recordedNot recorded2026-07-16 · model_announcementArtificial Analysis
GPT-5.6 Terra (max)0.389380530973451Not recordedNot recorded2026-07-09 · model_announcementArtificial Analysis
GPT-5.5 (xhigh)0.376843657817109Not recordedNot recorded2026-04-23 · model_announcementArtificial Analysis
GPT-5.6 Luna (max)0.358407079646018Not recordedNot recorded2026-07-09 · model_announcementArtificial Analysis
GLM-5.2 (max)0.337020648967552Not recordedNot recorded2026-06-16 · model_announcementArtificial Analysis
GPT-5.4 (xhigh)0.33259587020649Not recordedNot recorded2026-03-05 · model_announcementArtificial Analysis
Claude Opus 4.6 (Adaptive Reasoning, Max Effort)0.3303834808259587Not recordedNot recorded2026-02-05 · model_announcementArtificial Analysis
Gemini 3.1 Pro Preview0.320058997050148Not recordedNot recorded2026-02-19 · model_announcementArtificial Analysis
Kimi K2.60.284660766961652Not recordedNot recorded2026-04-20 · model_announcementArtificial Analysis
GPT-5.4 mini (xhigh)0.281710914454277Not recordedNot recorded2026-03-17 · model_announcementArtificial Analysis
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)0.28023598820059Not recordedNot recorded2026-02-17 · model_announcementArtificial Analysis
Gemini 3 Flash Preview (Reasoning)0.277286135693215Not recordedNot recorded2025-12-17 · model_announcementArtificial Analysis
GPT-5.4 nano (xhigh)0.249262536873156Not recordedNot recorded2026-03-17 · model_announcementArtificial Analysis
DeepSeek V4 Pro (Reasoning, Max Effort)0.242625368731563Not recordedNot recorded2026-04-24 · model_announcementArtificial Analysis
Qwen3.7 Plus0.224188790560472Not recordedNot recorded2026-06-01 · model_announcementArtificial Analysis
Grok 4.3 (high)0.170353982300885Not recordedNot recorded2026-04-30 · model_announcementArtificial Analysis
Qwen3.5 397B A17B (Reasoning)0.153392330383481Not recordedNot recorded2026-02-16 · model_announcementArtificial Analysis
Step 3.7 Flash0.148230088495575Not recordedNot recorded2026-05-29 · model_announcementArtificial Analysis
DeepSeek V3.2 (Reasoning)0.145280235988201Not recordedNot recorded2025-12-01 · model_announcementArtificial Analysis
GLM-5 (Reasoning)0.14454277286135694Not recordedNot recorded2026-02-11 · model_announcementArtificial Analysis
Grok 4.20 0309 (Reasoning)0.14233038348082597Not recordedNot recorded2026-03-10 · model_announcementArtificial Analysis
Gemini 3.1 Flash-Lite0.121681415929204Not recordedNot recorded2026-03-03 · model_announcementArtificial Analysis
Kimi K2.5 (Reasoning)0.11504424778761062Not recordedNot recorded2026-01-27 · model_announcementArtificial Analysis
MiniMax-M2.70.106194690265487Not recordedNot recorded2026-03-18 · model_announcementArtificial Analysis
gpt-oss-120b (high)0.0309734513274336Not recordedNot recorded2025-08-05 · model_announcementArtificial Analysis
MiMo-V2.5-Pro0.0243362831858407Not recordedNot recorded2026-04-22 · model_announcementArtificial Analysis
Nemotron 3 Super 120B A12B (Reasoning)0.0184365781710914Not recordedNot recorded2026-03-11 · model_announcementArtificial Analysis
gpt-oss-20b (high)0.00737463126843658Not recordedNot recorded2025-08-05 · model_announcementArtificial Analysis

Put this benchmark in context

Tools and agents benchmarksReasoning and math benchmarks