Resources · Benchmarks
Independent benchmarks.
Methodology, datasets, and head-to-head results. Reproducible, open where possible.
Benchmark
Indic-Eval v2: 22-language LLM benchmark
An open multilingual benchmark across 22 Indic languages.
May 20, 2026
Benchmark
Voice-AgentContainment Benchmark (VACB)
Standardised tasks + scoring for voice-agent containment.
Apr 26, 2026
Benchmark
Hallucination-in-Code: 12-language eval suite
Code generation hallucination eval across 12 languages.
Mar 30, 2026
Benchmark
Enterprise LLM Performance Benchmark
Comprehensive evaluation of enterprise LLMs across reasoning, accuracy, latency, and response consistency.
Mar 18, 2026
Ready to build
AI you can trust?
Talk to a solutions architect — get a pilot scoped in 48 hours.