Benchmarks
GPQA 2026
Official General-Purpose Question Answering benchmark for 2026.
MARKETS Benchmark
Nemo acer tabula creta clibanus claudeo causa exercitationem tabesco deduco.
PLATFORMS Benchmark
Demens absque stultus suus totus adinventitias tabella desidero creo vilis.
RELATIONSHIPS Benchmark
Conservo eos tui adipiscor auctus dens dedico decimus trepide.
SCHEMAS Benchmark
Valde aggredior civitas truculenter.
Recent Test Runs
Test Run #1 - MARKETS Benchmark
Models: Gemini 2.5 Pro, GPT O3 | Languages: de, es, it
RUNNING4 days ago
Test Run #2 - PLATFORMS Benchmark
Models: GPT O3, Llama 4, Claude 4 Sonnet | Languages: de, es, en, fr, it
RUNNING2 days ago
Test Run #3 - GPQA 2026
Models: Gemini 2.5 Pro, GPT O3, Llama 4, Claude 4 Sonnet | Languages: de, en
COMPLETED24 days ago
Test Run #4 - RELATIONSHIPS Benchmark
Models: Claude 4 Sonnet | Languages: fr, de, it, es, en
COMPLETED22 days ago
Test Run #5 - RELATIONSHIPS Benchmark
Models: Llama 4 | Languages: en, de, it, fr
COMPLETED4 days ago
Test Run #6 - SCHEMAS Benchmark
Models: Gemini 2.5 Pro, GPT O3 | Languages: en, de
RUNNING6 days ago
Test Run #7 - PLATFORMS Benchmark
Models: Claude 4 Sonnet, Llama 4 | Languages: fr, en, es, de, it
RUNNING2 days ago
Test Run #8 - SCHEMAS Benchmark
Models: Gemini 2.5 Pro, Claude 4 Sonnet, GPT O3, Llama 4 | Languages: it, fr
COMPLETED19 days ago
Test Run #9 - RELATIONSHIPS Benchmark
Models: GPT O3, Gemini 2.5 Pro, Llama 4, Claude 4 Sonnet | Languages: es, it, en
FAILED8 days ago
Test Run #10 - GPQA 2026
Models: Claude 4 Sonnet, Gemini 2.5 Pro, GPT O3 | Languages: fr, es, it, de
COMPLETED19 days ago