{"slug": "context-grapher-jedify-cuts-ai-token-costs-75-percent", "title": "Context grapher Jedify cuts AI token costs 75 percent", "summary": "Jedify, a context graph technology company, announced that its approach can cut AI token costs by up to 75 percent, citing research showing its context graph architecture averaged 25,036 raw tokens per SQL generation call with 87 percent accuracy, compared to 50,000–150,000 tokens and 60–70 percent accuracy for traditional schema-injection methods. The company's co-founder and CTO, Adi Elimelech, said that pre-encoding business logic into a context graph reduces the model's reasoning burden and improves token efficiency and accuracy.", "body_md": "# Context grapher Jedify cuts AI token costs 75 percent\n\n[Jedify](https://www.blocksandfiles.com/ai-ml/2026/06/10/autonomous-context-graphs-get-jedi-powers/5253395), which says AI needs augmenting with context graph technology, says its tech can cut AI token costs by up to three quarters.\n\nContext graphs store the relationships between entities (eg; people, products, locations) in some kind of information structure, such as a database. A [graph database](https://www.blocksandfiles.com/storage-management/2022/08/18/graph-database/1595670?_gl=1*o0azif*_ga*MzkxNDQyMTIwLjE3NzcwMzc0NTc.*_ga_NSDTXHMMN0*czE3ODc4MzkwMTQkbzMxNSRnMSR0MTc4Nzg0MDcwMSRqNjAkbDAkaDA.) is designed to store, manage, and query data as nodes (entities) and edges (relationships) in a graph structure. Unlike relational databases, which use tables and joins, graph databases prioritize relationships, enabling efficient traversal and analysis of complex, interconnected data. A knowledge graph is a graph database used for AI and a context graph is used in the same area but reveals which relionships are valid.\n\nAdi Elimelech, co-founder and CTO of [Jedify](https://jedify.com/), said: “Most of the industry treats token-efficiency and cost-efficiency as the same problem, but they're not. When business logic is pre-encoded into a context graph, the model isn't reasoning from scratch on every query. That's the mechanism that matters, regardless of which vendor's implementation you're looking at.\"\n\nJedify says rather than exposing AI to your entire database schema on every query, a context graph pre-encodes your business logic, including definitions, calculations, relationships and rules, into a structured knowledge base. That context is continuously maintained and autonomously expanded as the business evolves, allowing the graph to grow alongside new data, relationships and business logic. When a question is asked, the system pulls only the handful of entities relevant to that specific question so it can deliver a more compact but more precise package to the model.\n\nThe company’s research tested its context graph approach against a live production data warehouse, measuring 100 business questions across three complexity tiers, each run twice for a total of 200 graded data points. The context graph architecture averaged 25,036 raw tokens per SQL generation call and answered 87 percent of graded runs correctly. For comparison, published research on traditional approaches that inject raw database schema into an LLM reports 50,000 to 150,000 tokens per call and 60 percent to 70 percent accuracy, while multi-agent systems built on schema injection, such as CHESS, report figures near 339,965 tokens per request. Those baseline figures come from other studies run on different schemas and question sets, not from a same-warehouse comparison.\n\nUsing a token ROI framework, which weighs both cost and answer accuracy, the research found that the context graph approach delivered a 4.6X higher return than schema-injection approaches and 20X higher than naive multi-agent pipelines. Schema-injection approaches returned correct answers roughly 60 percent to 70 percent of the time in the study, compared with near-perfect accuracy for the context graph approach on entities it covers.\n\nAt 100-table enterprise scale, the context graph architecture tested in the study injected roughly 50 percent fewer tokens per SQL call than a raw schema baseline. At 200 tables, that gap exceeded 75 percent.\n\nElimelech said: \"There's a data ownership piece to this, too. When your business logic lives in a graph you control, you're not handing a frontier model provider your entire schema and every join and filter rule on every single call.\"\n\nBecause a context graph can narrow each query down to only the entities it needs before SQL generation, the reasoning burden on the underlying model shrinks. The study's production routing analysis suggests that roughly 85 percent of enterprise analytics queries can run on lower-cost, open-source models rather than high-end frontier models, without a meaningful drop in accuracy. That means you can use lower-cost infrastructure.\n\nDownload a white paper discussing all this [here](https://jedify.com/context-graph-advantage/?utm_source=globenewswire&utm_medium=pressrelease&utm_campaign=#c-download).", "url": "https://wpnews.pro/news/context-grapher-jedify-cuts-ai-token-costs-75-percent", "canonical_source": "https://www.blocksandfiles.com/ai-ml/2026/08/27/context-grapher-jedify-cuts-ai-token-costs-75-percent/5293003", "published_at": "2026-08-27 15:06:08+00:00", "updated_at": "2026-08-27 15:20:32.175952+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-infrastructure", "ai-research"], "entities": ["Jedify", "Adi Elimelech", "CHESS"], "alternates": {"html": "https://wpnews.pro/news/context-grapher-jedify-cuts-ai-token-costs-75-percent", "markdown": "https://wpnews.pro/news/context-grapher-jedify-cuts-ai-token-costs-75-percent.md", "text": "https://wpnews.pro/news/context-grapher-jedify-cuts-ai-token-costs-75-percent.txt", "jsonld": "https://wpnews.pro/news/context-grapher-jedify-cuts-ai-token-costs-75-percent.jsonld"}}