Finviz alternative · free

DeepVane: the Finviz alternative that publishes its math

Finviz tells you which stocks crossed which threshold. DeepVane tells you which stocks score 0–100 across twelve academic factors, how confident the model is in each call, and which patterns are firing right now — with every formula on a public methodology page.

Try NVDA →Open methodology

DeepVane vs Finviz

Both are free. Different mental models.

FeatureDeepVaneFinviz
Free, no credit cardYesYes
Universe size374 curated tickers~9,000 US tickers
Multi-factor academic score12 factor families, 0–100No — filter-based
Confidence interval per signalMondrian conformal intervalsNo
Regime detectionBayesian (BOCPD on SPY)No
Phase 3 trial failure predictionBayesian, every biotech tickerNo
Pattern library with academic source18 patterns, every cite publicTechnical patterns, no sources
Open methodology page/methodology + /diag/engineProprietary
Public track recordDay-30 returns from 2026-05-16None published
Real-time price snapshots5-min cron during market hoursReal-time
Heatmaps + sector viewsSector rotation pageBest-in-class heatmap
News aggregationPer-ticker, scored by qualityBest-in-class flow

What you get on DeepVane

Twelve factors, not seventy filters

Finviz screens by individual technical / fundamental thresholds. DeepVane combines twelve peer-reviewed factor families (quality, value, momentum, PEAD, insider flow, NLP tone, options, accruals, short interest, sector momentum, spillover, factor interaction) into one Bayesian-shrunk composite score. Picking with one number that has academic provenance is faster than picking with twenty filters that don't.

A confidence interval on every signal

Every DeepVane score ships with a Mondrian conformal prediction interval — the model says "score 78, but expect ± 14 with 90% coverage on this regime / sector cohort." A finviz score is just a number; it doesn't tell you when to ignore itself. That matters when you size positions.

The math is on a public page

Every factor cites the paper it implements (Novy-Marx 2013 quality, Loughran-McDonald 2011 NLP, Hay 2014 Phase 3 base rate, Asquith 2005 short squeeze). The engine's 21 mathematical invariants run live at /diag/engine — auditable on every request. We trade some opacity for being verifiable instead of credible-by-assertion.

Phase 3 trial failure prediction

No equity screener does this. DeepVane runs a Bayesian model on every Healthcare-sector ticker: anchor on Hay et al. 2014 base rates, update with log-likelihood-ratios from enrollment velocity, endpoint amendments, mechanism risk, and cash-burn runway. The output is a colour-coded PH3 pill on every biotech card.

Common questions

Is DeepVane really free, like Finviz?

Yes during the early-access window. All twelve factor scores, all eighteen confluence patterns, all Phase 3 failure predictions, and all conformal prediction intervals are publicly readable without a paywall. No credit card. The Pro tier launches on 16 May 2026 once the first 30 days of forward returns calibrate the model — current beta usage stays free.

Why does DeepVane only cover 374 tickers vs Finviz's 9,000?

Two reasons. (1) The factor pipeline reads SEC EDGAR 10-K / 20-F filings, ClinicalTrials.gov, PubMed, options and short-interest feeds — running this on 9,000 names exceeds free-tier API quotas. (2) The 374 are curated for liquidity and coverage of the academic factor literature. We expand as data quality allows; the universe was 234 in March 2026 and grew by ~40% in one cycle.

Can I use DeepVane for technical screening like Finviz?

No. DeepVane is a fundamental + factor + Bayesian-regime engine, not a chart-pattern screener. If you need RSI / MACD / 52-week-high filters, Finviz remains the fastest tool for that. If you want to know which stocks score highly on a peer-reviewed multi-factor model with honest uncertainty intervals, DeepVane is the answer.

How is DeepVane's methodology different from Finviz's screening?

Finviz is a filter-based screener — the user chooses the criteria. DeepVane is a model — the engine combines factors into one composite under regime-conditional weights, surfaces confluence patterns, and outputs a 0–100 score with a confidence interval. The closest analogy is "Finviz is a spreadsheet, DeepVane is a forecast."

Try a ticker in 5 seconds

No sign-up wall. Type any of 374 tickers, see the full 12-factor APEX breakdown, the regime context, every firing pattern, and the conformal prediction interval. Then compare with Finviz — that's the cleanest test.

Try a ticker →See track record