# Bybit vs dYdX (slim)

> dYdX scores 69.5 (B) to Bybit's 57.3 (C) for exchange market data. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/bybit-vs-dydx.md (~4,600 tokens) · this version ~2,430 tokens · JSON https://www.anchorterminal.com/compare/bybit-vs-dydx.json · canonical https://www.anchorterminal.com/compare/bybit-vs-dydx
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-03
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

dYdX scores 69.5 (B) on agent readiness against Bybit's 57.3 (C), and leads in 4 of 7 scored categories. Bybit leads on schema & documentation. Both do exchange market data.

- Bybit: C 57.3, rank #745 of 1349 · https://www.anchorterminal.com/tools/bybit.min.md
- dYdX: B 69.5, rank #221 of 1349 · https://www.anchorterminal.com/tools/dydx.min.md
- Bybit, good for An agent trading spot, perpetuals, futures and options in a person's Bybit account outside the excluded jurisdictions, with a demo environment to rehearse on. Ahead on Schema & documentation, 73 against 58. Also Runs on your own machine.
- dYdX, good for An agent trading perpetuals from its own Cosmos wallet or through a permissioned key on an owner's account, outside the US, Canada and UK. Ahead on Reliability, 83 against 45; Security & auth, 78 against 60; Payments & pricing, 60 against 40; Maintenance & community, 83 against 66. Also Open source.

| Category | Bybit | dYdX |
| --- | --- | --- |
| Reliability (16%) | 45 | 83 |
| Performance (10%) | pending | pending |
| Schema & documentation (13%) | 73 | 58 |
| Agent ergonomics (13%) | 66 | 66 |
| Security & auth (14%) | 60 | 78 |
| Payments & pricing (10%) | 40 | 60 |
| Task success (10%) | pending | pending |
| Maintenance & community (7%) | 66 | 83 |
| Transparency & trust (7%) | 51 | 50 |

| Fact (where they differ) | Bybit | dYdX |
| --- | --- | --- |
| Vendor | Bybit | dYdX Trading Inc. |
| Transports | stdio |  |
| Auth | API key | OAuth or key |
| Licence | none | The clients are AGPL-3.0 with a use condition tied to the v4 Terms of Use; the chain software (dydxprotocol/v4-chain) carries a custom licence GitHub cannot classify. The hosted indexer at indexer.dydx.trade and the front end are run by third parties, not dYdX Trading Inc. |
| Tools exposed | 398 | 0 |
| Read-only variant documented | yes | no |
| llms.txt | no | yes |
| Last release | 2026-10-30 | 2026-10-29 |
| Terms last updated | couldn't be read | 2026-03-12 |
| Privacy policy last updated | couldn't be read | 2025-06-06 |
| Customer content may train models | couldn't be read | not found in the text |
| Terms restrict automated access | couldn't be read | not found in the text |
| Terms restrict benchmarking | couldn't be read | not found in the text |
| Terms or service can change without notice | couldn't be read | yes |
| Arbitration or class-action waiver | couldn't be read | yes |
| Popularity | 43 stars | 119 stars, 86k npm/wk |

Why the edge:
- Reliability, dYdX +38: Bybit 45 (Graded with the hosted lines on the V5 API at api.bybit.com. No status page found (status.bybit.com does not resolve, the docs link none), so 0 for the page and 5 for no readable history. Rate limits are published per endpoint and per UID with numbers, plus a 600 per 5 second IP limit (15). Every response carries X-Bapi-Limit, X-Bapi-Limit-Status and X-Bapi-Limit-Reset-Timestamp, retCode 10006 and HTTP 403 are documented with a wait rule, and orders take a unique orderLinkId that amend and cancel accept (15). No SLA (0). V5 is the current production API, not a beta (10). 45.); dYdX 83 (Read with the hosted lines on the indexer API and the chain endpoints the clients submit to. status.dydx.trade is an Atlassian Statuspage with Frontends, dYdX Chain, Indexer, Order Gateway and Client Libraries components and an RSS history (20). From 1 August to 3 November 2026 the feed has eight items, seven scheduled maintenance windows of 29 minutes to 2 hours (indexer on 9 September, 8 and 15 October, protocol v9.7 on 11 September and v9.8.0 on 12 October, WebSocket compression on 26 and 28 October) and one degraded instant-deposit path through an onboarding partner from 24 to 29 September, outside the trading API. Clean or trivial (30). Rate limits documented with numbers for the indexer, the block limits and the public nodes (15). No 429 shape or Retry-After for the indexer; short-term orders expire by `goodTilBlock` and client IDs let an agent find an order after a timeout (8 of 15). No SLA (0). Mainnet generally available; the Order Entry Gateway is marked testnet-only (10). 20 + 30 + 15 + 8 + 0 + 10 = 83.)
- Payments & pricing, dYdX +20: Bybit 40 (No machine payment protocol (0). The fee schedule is public without login, with VIP 0 maker and taker rates per product and the tier criteria (20). Reading market data needs no key and the demo environment costs nothing, no card is asked for at sign-up (20). A person must register, pass identity verification and mint the key in a browser (0). 40.); dYdX 60 (No machine payment protocol (0). Fees public per tier without login (20). Testnet with a faucet and no card anywhere (20). No signup; an agent with a funded wallet trades, subject to the geographic terms (20). Total 60.)
- Security & auth, dYdX +18: Bybit 60 (API keys scoped by product group with a readOnly flag, a separate Withdraw value, IP binding, a 90-day expiry for unbound keys, HMAC or RSA signing, and revocation on the management page; no OAuth (24 of 30). A read-only key and a withdrawal-free key are first-class, but the MCP server has no read-only mode and relies on the client's approval prompt for writes (14 of 20). Tools return the exchange's own data, not third-party text (10). The API key info endpoint shows permissions, IPs and expiry, and an IP change log exists for institutional accounts, but no per-call audit log was found (7 of 15). No security.txt on bybit.com or the docs host, no disclosure policy or bug bounty found on a page we could read, proof-of-reserves articles in the help centre (5 of 20). 60.); dYdX 78 (Orders are signed by the account key or by a permissioned key the owner adds as an authenticator, limited by message type, subaccount and market with AllOf or AnyOf, and revoked by `remove_authenticator`; no shared secret and nothing in a URL (28 of 30). A key can be confined to placing and cancelling orders on one subaccount with withdrawals and transfers excluded, though not to a position or order size, and the interface's default API wallet trades all cross-margin pairs (16 of 20). The API returns no untrusted content (10). Every action is a chain transaction and the indexer lists fills, orders and transfers per subaccount, and authenticators can be listed; no per-key log (10 of 15). A Cantina bug bounty up to $1,000,000 sponsored by dYdX Trading Inc. over the protocol, indexer, web client and SDKs, an Informal Systems audit; no security.txt and no certification (14 of 20). Total 78.)
- Maintenance & community, dYdX +17: Bybit 66 (Read as a hosted service with an open-source MCP server. The MCP package shipped 2.1.26 on 30 October 2026 and the V5 changelog's latest entry is 29 October (30). 31 npm versions since 16 April 2026, 11 of them in the last 90 days, and 24 changelog entries in the same window (20). The trading-mcp repository has 4 open issues and 18 forks; whether issues get answers was not checked, and the API has no public forum we could read (10 of 25). Not in the official MCP registry (0). The package has four dependencies, no install scripts, a publish workflow and no test workflow (6 of 10). 66.); dYdX 83 (Read on the clients and the chain. v4-clients was pushed on 29 October 2026 and protocol v9.8.0 shipped on 12 October (30). Protocol v9.7 on 11 September, v9.8.0 on 12 October, v4-client-js tags 3.5.0 and 3.6.0, three or more in 90 days (20). A Discord, a help centre and 36 open issues on v4-clients whose replies couldn't be read (12 of 25). Official SDKs current, the TypeScript client at 3.6.0 and the Python client at 1.1.6 from February 2026 (12 of 15). CI workflows for lint, test, publish and audit, with trusted publishing and provenance on npm (9 of 10). Total 83.)
- Schema & documentation, Bybit +15: Bybit 73 (No OpenAPI or other machine-readable contract for the REST API; the MCP server's 398 tools have zod schemas turned into JSON Schema, so the contract exists for the MCP surface only (12 of 25). No llms.txt; the docs are a public Docusaurus repository (bybit-exchange/docs) whose MDX source an agent can read (5 of 10). The MCP tool descriptions state purpose, rules and an agent hint per tool, and the API reference gives one page per endpoint (15 of 20). Parameters are typed with enums such as category and orderType and required flags (12 of 15). Request and response examples on each endpoint page and an error code page of about 750 codes with text (14 of 15). A dated changelog with 24 entries in the last 90 days, versioned V3 to V5 (15). 73.); dYdX 58 (No OpenAPI file found for the indexer; the docs publish a types reference of about 200 pages (OrderSide, TimeInForce, OrderStatus and the rest) and the chain's protobuf messages are in the clients (8 of 25). llms.txt and Markdown docs (10). Endpoint pages describe each call and the guides say when to use short-term against long-term orders and the indexer against the node, but descriptions are short (12 of 20). Enums and types on every field in the reference, with pagination and cursor parameters typed (11 of 15). Examples in Python, TypeScript and Rust for each step; errors are described by HTTP status (400 malformed, 404 not found) with no code list (9 of 15). A /v4 path, protocol releases dated on the status page, deprecated rate-limit fields marked, but no API changelog (8 of 15). Total 58.)