Ecosystem All ecosystem services Screener.in ↗

Screener.in — the human fundamentals desk

Gold standard for fundamentals viewing on Indian markets. Deeper history than anyone, focused on fundamentals.

Decision matrix

What Screener.in is great at

• Fundamentals viewer. Quarterly + annual data, peer comparison, ratios, charts — rendered for humans.
• Custom screens. Write a formula in their DSL, save and share screens.
• Long history. Backfill goes deep into the PDF era (back to 1997 in many cases).
• Premium tier. Unlocks more screens, faster refresh, peer-comparison limits.
• API status. No public developer API. No MCP. Closed surface.
• What it does not ship. Programmatic access, MCP, MF screener at our depth, concall transcripts as a corpus.

Where Genka complements Screener

• JSON API + MCP. Same data shape humans see on Screener, but addressable from code with X-API-Key auth and a self-describing tool catalog.
• Source-cited responses. Every numeric field carries a primary source in meta.sources[]. Where we cross-validate XBRL vs Screener, conflicts surface as bs_source: both_disagree with per-column diff_pct.
• Concall corpus. ~23,000 transcripts; BM25 + grounded LLM Q&A.
• MF analytics. 4,700 funds, SEBI taxonomy, 20y NAV, screener, backtester, holdings overlap.
• Indicator panel. Pre-computed RSI, MACD, DMA, multi-window CA-adjusted returns.

Compose them

Screener for the human eye, Genka for the agent loop:

# Pattern: Your dashboard renders the agent's findings,
# and links each company to its Screener page for the human's deep read.

# 1. Agent screens via Genka MCP:
"Find Nifty 500 names where ROE has expanded for 4 consecutive
 quarters and operating margin is above 5y-median."

# 2. Render each hit with two links:
GENKA_DETAILS = f"https://genka.dev/v1/companies/{sym}/ratios"
SCREENER_PAGE = f"https://www.screener.in/company/{sym}/"
#                              ↑ for the human

Note on data lineage

Where Screener has historical data we can't get from primary XBRL (pre-FY22 balance-sheet totals are PDF-only at NSE), we backfill from Screener with attribution. Each affected response shows screener.in in meta.sources[]. We are not a Screener competitor on the consumer surface; we are the API for code that Screener doesn't ship.

Relevant Genka endpoints

• GET /v1/companies/{symbol}/ratios — quarterly fundamentals time series
• GET /v1/companies/{symbol}/balance-sheets — full BS, with source per row
• GET /v1/companies/{symbol}/coverage — which metrics/periods are populated
• POST /v1/screener — programmatic Nifty 500 filter
• POST /v1/concalls/answer — grounded Q&A
• GET /v1/mf/screener — mutual fund screener

Full catalog: /llms.txt

Working with Screener.in — what's available

• Custom-formula DSL. Screener's screen DSL is the killer feature. Syntax like Sales growth 5Years > 15 AND Debt to equity < 0.5 works for most ratios; full operator list is in their docs. Saved screens are shareable by URL.
• Pre-FY17 history is PDF. Screener's depth comes partly from parsing annual-report PDFs going back to ~1997. They're the gold standard for human reading; we backfill from them for pre-FY22 BS totals where NSE XBRL is unavailable.
• Premium tier. ₹3,500/yr unlocks more saved screens, custom ratios, and faster refresh. Useful if your workflow is human-in-browser; doesn't change the API picture (still none public).
• Concalls are not in Screener. Screener focuses on financials. For concall transcripts you need either each company's investor-relations site or an aggregator (we ship 23k via /concalls/search).