Comparator
Broker API Comparator
Alpaca vs IBKR vs Tradier vs Schwab vs Robinhood — auth, rate limits, order types, market data, MCP, fees. Pick the right broker before coding. Free.
- Inputs
- Scenario form
- Runtime
- Instant
- Privacy
- Client-side · no upload
- API key
- Not required
- Methodology
- Open →
Best fit
Alpaca
1/1 fit-score · $0 (PFOF on equities) · auth complexity 1/5 · 5 of 5 brokers match
Solo AI operator, paper-to-live parity
Ranked brokers
Solo AI operator, paper-to-live parity
Fit-score
1/1
Auth
API Key + Secret complexity 1/5
Commission
$0 (PFOF on equities)
Rate limits
200 req/min (free), higher on paid tiers
MCP
Official Official MCP V2, 61 actions, idempotent order submission
Asset classes
Order types
Options-first retail, multi-leg strategies
Fit-score
1/1
Auth
OAuth 2.0 complexity 3/5
Commission
$0 equities (PFOF), $0.35/contract options Pro
Rate limits
120 req/min market data, 60 req/min trading
MCP
Community Community MCP; idempotency is a manual patch
Asset classes
Order types
Established retail, post-TDA API migration
Fit-score
1/1
Auth
OAuth 2.0 (3-legged) complexity 4/5
Commission
$0 equities, $0.65/contract options
Rate limits
120 req/min per endpoint (published tier)
MCP
None No official or listed community MCP server
Asset classes
Order types
Casual retail; not recommended for agents
Fit-score
1/1
Auth
OAuth 2.0 (unofficial endpoints) complexity 4/5
Commission
$0 (PFOF), $5/mo Gold for margin + extras
Rate limits
Not publicly documented; low throughput
MCP
None No MCP; unofficial SDKs only
Asset classes
Order types
Multi-asset-class, international, scale
Fit-score
1/1
Auth
OAuth + TWS/IB Gateway complexity 5/5
Commission
Tiered per-share or fixed; ~$0.0035/sh tiered
Rate limits
~50 msg/sec per connection; 50 orders/sec burst
MCP
Community Community CLI MCP; audit before production
Asset classes
Order types
How this ranking works
Each broker is scored against the filters you've enabled. Fit-score is an unweighted count of matching criteria. Brokers that fail any required criterion are pushed below qualifying ones and dimmed.
Auth complexity is subjective — 1 is a plain API-key header, 5 means a persistent gateway process on your machine. Rate limits shown are vendor-published tier rates; real-world limits may be lower under concurrent load. See the methodology for sources and caveats.
Data last checked 2026-04-22. Broker API terms drift; always verify on the broker's developer portal before committing.
How to use
Step-by-step
- 1
Filter brokers by required asset coverage (US equity, options, futures, crypto).
- 2
Read the rate-limit table and the latency benchmarks (sandbox p50/p95/p99).
- 3
Read commission cost (where applicable) plus PFOF analysis (where 'free').
- 4
Read the auth method and security posture. Some brokers require OAuth; others use API key.
- 5
Test in paper-trading mode before committing real capital. Sandbox latency is best-case; production tail latency varies with deployment region and broker traffic.
For agents
Use in an agent
Same math, same result shape as the UI above — as a static ES module. No HTTP request, no auth, no rate limit.
import { compute } from "https://aifinhub.io/engines/broker-api-comparator.js";Contract: /contracts/broker-api-comparator.json Full agent guide →
Glossary references
Terms used by this tool
Questions people ask next
FAQ
Which brokers does it cover?
Documented on the methodology page: Alpaca, Interactive Brokers, Tradier, Tradestation, Robinhood, Coinbase Advanced Trade, Kraken Pro. Each is graded on rate limits, asset coverage, paper trading, latency, and authentication style. The list is updated quarterly.
How are latency numbers measured?
Synthetic order placement and cancellation against the broker's sandbox/paper environment, measured at p50/p95/p99 across 10K samples. Methodology page shows the test client and the cloud regions used. Real-world production latency depends heavily on your colocation — the numbers are best-case sandbox latency, not promises.
Does it model commission cost?
Yes for the brokers that charge commissions (IBKR fixed, IBKR tiered, Tradestation flat). For zero-commission brokers (Alpaca, Robinhood) the tool flags PFOF (payment for order flow) as an effective cost. The methodology page documents how PFOF translates into worse fills.
Why isn't broker X listed?
Either the broker doesn't expose a documented public API (many full-service brokers), or our coverage assessment hasn't completed yet. Methodology page lists the criteria for inclusion: documented API, sandbox available, asset coverage breadth, and reliability track record.
Should I trust 'free' brokers?
Free brokerage is paid for somewhere — typically PFOF on equity orders or wider spreads on options. The tool surfaces this. For low-volume retail, the difference is usually a few dollars per year — fine. For high-volume strategies, the cost can dominate.
Related deep dive
All articles →Read further
Long-form context behind the tool output.
- Pillar · Guide·7 min
Interactive Brokers API Updates 2026: TWS 10.47 Changelog
Interactive Brokers API updates 2026: TWS 10.47 (May 20) removes reqFundamentalData, adds the $LEDGER- per-currency prefix; 10.45 stable. Verified changelog.
Read - Comparison · Benchmark·10 min
Choosing a Broker API 2026: Rate Limits, Fees, Auth
Choosing a broker API 2026 — Alpaca vs IBKR vs Tradier vs Schwab vs Robinhood on the axes that bite: auth, order types, rate limits, and fees.
Read - Comparison · Benchmark·11 min
Alpaca vs Tradier: Options + MCP Coverage Compared
Comparator returns Alpaca as the only strict-filter fit (4/4); Tradier ties when auth relaxes. MCP directory confirms one official broker MCP.
Read
Complementary tools
Users of this tool often explore
Data-Vendor TCO Calculator
Compute annual cost of market data across Databento, Polygon, Alpaca, Tiingo, FMP, and Alpha Vantage for your exact universe, bar resolution, history.
Kalshi vs Polymarket Arb Scanner
Daily-refreshed scan of arbitrage candidates across Kalshi and Polymarket. Paired contract matching, tax + resolution-risk overlay, no signup. Edge data.
Model Selector for Finance
Input task, latency budget, cost budget, context size, and quality sensitivity; get ranked model recommendations with rationale — grounded in published.