Intelligence

Focus areas

• Model capabilities

  4 entities · 6 metrics with public observations

• Inference economics

  1 entities · 2 metrics with public observations

• AI infrastructure

  3 entities · 5 metrics with public observations

Entities

• Claude Opus 4.6model · 1 observations
• Claude Opus 4.8model · 4 observations
• DeepSeek-V4-Flashmodel · 2 observations
• DeepSeek-V4-Promodel · 5 observations
• SAM 3.1 Object Multiplexmodel · 1 observations

Metrics

• API list price — inputapi_list_price_input_usd_per_mtok · 2 observations · 1 entities
• API list price — outputapi_list_price_output_usd_per_mtok · 2 observations · 1 entities
• CorpusQA 1M accuracycorpusqa_1m_acc · 1 observations · 1 entities
• MRCR 1M mean match ratiomrcr_1m_mmr · 3 observations · 2 entities
• Relative single-token inference FLOPs vs baselinerelative_inference_flops_vs_baseline · 2 observations · 2 entities
• Relative accumulated KV cache vs baselinerelative_kv_cache_vs_baseline · 2 observations · 2 entities
• Relative throughput vs prior releaserelative_throughput_vs_prior_release · 1 observations · 1 entities

Context only

Entities with evaluation context or metric definitions in the pilot, but no public numeric observation yet.

• Vertex AI Gen AI evaluation serviceproduct · Definition / evaluation context only — no observation yet