Skip to content

FUTURE-002: Kalshi Blocked/Unimplemented Endpoints

Status: Tracking Created: 2026-01-16 Revisit: Q2 2026 (Subaccounts), Periodically (others)


Summary

This document tracks Kalshi API endpoints that are not implemented in our client, along with the reasons and conditions under which we would implement them.

Current coverage: 50/74 endpoints (68%)


Endpoints to Revisit (May Become Available)

1. Subaccounts (4 endpoints) - Revisit Q2 2026

Added to Kalshi API: Jan 9, 2026 (per changelog)

Endpoint Observed (2026-01-15)
POST /portfolio/subaccounts 404 in demo + prod
GET /portfolio/subaccounts/balances 403 in prod ("not available in production")
POST /portfolio/subaccounts/transfer 404 in demo + prod
GET /portfolio/subaccounts/transfers 403 in prod ("not available in production")

Why blocked: Feature is brand new (7 days old when we tested). Likely in phased rollout / account-permission gated.

Value if available: - Track "Macro Thesis" vs "Crypto Thesis" separately - Know which strategies are actually profitable - Proper capital allocation across approaches

Action items: - [ ] Re-check in Q2 2026 - [ ] Contact Kalshi support to request access - [ ] Implement when API returns real responses

2. Forecast Percentile History (1 endpoint) - Check Periodically

Added to Kalshi API: Sep 11, 2025

Endpoint Observed (2026-01-15)
GET /series/{series_ticker}/events/{ticker}/forecast_percentile_history Unverified 200 payload (see notes)

SSOT (OpenAPI): This endpoint requires auth and requires query params: percentiles, start_ts, end_ts, period_interval. Without valid parameters, Kalshi will return 400 bad_request.

Changelog vs OpenAPI: The Sep 11, 2025 changelog entry calls this GET /forecast_percentiles_history (no series/event path). The current OpenAPI defines the series/event-scoped path above.

Why blocked/unimplemented: We have not recorded a stable 200 response fixture yet. Until we can, treat this as unimplemented.

Value if available: - Historical forecast accuracy for calibration research - Know when to fade vs follow the crowd - Calibration edge at extremes

Action items: - [ ] Re-check periodically - [ ] Try different event types when available


Intentionally Not Implemented (Institutional/Security)

These endpoints exist in the OpenAPI but are intentionally not implemented for our use case.

3. RFQ/Communications (11 endpoints) - Institutional Only

Endpoint Reason
GET /communications/id Large block trades (1000+ contracts)
POST /communications/rfqs Institutional use
GET /communications/rfqs Institutional use
GET /communications/rfqs/{rfq_id} Institutional use
DELETE /communications/rfqs/{rfq_id} Institutional use
POST /communications/quotes Institutional use
GET /communications/quotes Institutional use
GET /communications/quotes/{quote_id} Institutional use
DELETE /communications/quotes/{quote_id} Institutional use
PUT /communications/quotes/{quote_id}/accept Institutional use
PUT /communications/quotes/{quote_id}/confirm Institutional use

Would implement if: Trading very large positions where orderbook liquidity is insufficient.

4. FCM (2 endpoints) - Explicitly Institutional

Endpoint Reason
GET /fcm/orders FCM members only
GET /fcm/positions FCM members only

Per changelog: "This endpoint requires FCM member access level... only intended for use by FCM members (rare)"

Would implement if: Building a brokerage integration.

5. API Keys (4 endpoints) - Security Risk

Endpoint Reason
GET /api_keys Security risk
POST /api_keys Security risk
POST /api_keys/generate Security risk
DELETE /api_keys/{api_key} Security risk

Why not: Programmatic API key management is a security risk. Better managed via Kalshi web UI with 2FA.

Would implement if: Building a multi-user service that needs to manage keys programmatically.

6. Multivariate Low-Value (2 endpoints) - Low Priority

Endpoint Reason
POST /multivariate_event_collections/{collection_ticker} Create market in collection (advanced/rarely needed)
GET /multivariate_event_collections/{collection_ticker}/lookup Lookup history (requires lookback_seconds, niche use)

Would implement if: Need to create combo markets or analyze historical lookups.


Key Insight: Tiers ≠ Feature Access

Kalshi's Basic/Advanced/Premier/Prime tiers control rate limits (request throughput):

Tier Read/sec Write/sec
Basic 20 10
Advanced 30 30
Premier 100 100
Prime 400 400

Rate limits are not a guarantee of endpoint access. Some endpoints/features are permissioned by API usage level (e.g., OpenAPI restricts creating API keys with user-provided RSA keys to Premier/Market Maker usage levels).


Cross-References

Item Relationship
DEBT-015 Detailed endpoint analysis
kalshi-openapi-coverage.md Coverage tracking
kalshi-api-reference.md Vendor docs with access categories