Novacore is the Novamine API (Application Programming Interface) and chat: a drop-in step at the point where an existing workflow needs non-consensus breadth.
Ask one question and get a diverse set of responses, all highly usable, where a standard AI model repeats the same two or three. Ranked first on Carnegie Mellon’s NoveltyBench for both novelty and utility.
Your data is never used for training. Every response is scored for novelty and utility before you see it.
Independently reproduced and merged by the Carnegie Mellon NoveltyBench maintainers (10 September 2026). Every benchmark figure on this page was scored by the benchmark's own scoring.
Until now you had to choose. Push an AI model for variety and the answers get stranger but less dependable. Hold it to quality and it falls back to the answer everyone already has.
Novamine removes that trade-off. On the independent benchmark it produced the most different answers of any system, and those answers were more useful, not less.
NoveltyBench is an independent test of whether an AI system can give answers that are both different and good. Novamine's entry was reproduced and merged by the benchmark's maintainers on 10 September 2026. It ranks first for novelty and first for utility, and it is the only system to hold both. See the leaderboard.
The NeurIPS 2025 best paper showed that AI models give near-identical answers to open questions. We repeated the study using the author's own test, then applied Novamine. The share of questions stuck in the “hivemind” fell from 96% to 1%, and the effect carried over to a model from a second AI company.
The reply from the study's author:
Novacore is a drop-in step: 3 to 60 minutes, Rigour 0 to 3, at the point where an existing workflow needs non-consensus breadth. Your system stays as it is before and after the call. Six places it is used: three everyday workflows and three high-throughput uses of the same API.
A consultant puts a client problem to your internal LLM assistant.
A mechanism, material or process problem, with its design constraints fixed.
Market entry, growth, pricing or capital allocation.
A forecast, decision or debate question goes to your agent ensemble.
A model to tune, or an evaluation set to build.
A question put to your retrieval-augmented assistant.
One high-probability answer: competent, and the one every firm on the same model receives.
In-house models and literature-led ideation converge on the known mechanisms.
Analyst work and standard AI return the plays your competitors are already modelling.
Agents built on one model reason alike, so their errors add up instead of cancelling.
Synthetic data from one model clusters around the same few answers.
Your retrieval layer pulls the context from your knowledge base, data lake or warehouse.
The user asks for alternatives; the assistant routes the brief to Novacore.
High rigour for a few defensible mechanisms, or a wide pool for your own models to triage.
The team calls for non-consensus routes, tested to the rigour the decision needs.
Your orchestrator asks Novacore for independent lines of reasoning.
Your pipeline requests scored, distinct responses for each prompt.
Question and context go to Novacore, in place of the single consensus answer.
Non-overlapping answers returned in-thread, each scored for novelty and utility.
Mechanistically distinct, screened against prior art and scored.
Distinct strategies, each challenged and scored against the brief.
Distinct, scored reasoning paths: one for each agent, debater or forecaster.
Responses from across the range, each scored for novelty and utility.
Distinct, scored readings of the same evidence.
The user selects a direction and iterates in your assistant, context intact.
Your in-silico triage and wet-lab validation proceed unchanged.
Your team runs the financials on the strongest and takes one to the board.
Your mixture-of-agents, debate or forecast step runs unchanged. crowd error = average error − diversityDiversity Prediction Theorem: Page, The Difference, 2007
Your next model learns from the full range, not the collapsed middle.
Mode collapse does not follow you into your own data.