Synapse Review

How Synapse scores in bridges

This entry rates the Synapse Bridge in our bridge ranking. The bridge routes each transfer through liquidity pools (via nUSD and nETH intermediate tokens), Circle's Cross-Chain Transfer Protocol (CCTP) for USDC, or the Synapse Intent Network, a relayer request-for-quote (RFQ) system. The basis is desk research of the Synapse documentation and the synapsecns GitHub repositories, checked on November 6, 2026. No transfer was made. The category uses five criteria, each marked 0 to 10 and weighted equally, as set out in our ratings methodology.

The overall score is the plain average: (2 + 6 + 9 + 7 + 5) / 5 = 29 / 5 = 5.8.

Three marks changed at this check, and the overall rose from 3.4 to 5.8. The earlier review could not render the documentation and found a 404 on the chains page. Both now load. Trust model disclosure rose from 2 to 6 for the Intent Network security pages. Chain coverage rose from 2 to 7 for the 19 networks on the Chains & Tokens page. Audit disclosure rose from 2 to 5 for the named reports in the sanguine repository.

Strongest criteria: licence and chain coverage

Licence & code openness scores 9. MIT is a recognised open-source licence, which the rubric places at the top. The mark stays one point short because the repository holding the pool-bridge contracts has no LICENSE file and relies on per-file headers. The category applies the same one-point deduction to Wormhole, whose Apache 2.0 licence is a short-form notice.

Chain coverage scores 7. The list is specific, with tokens per network, and includes newer EVM chains such as Unichain, World Chain, HyperEVM and Berachain. It is narrower than the lists of the leading peers, and the Bridge page's own "20+ EVM and non-EVM blockchains" does not match the 19 enumerated. The mark follows the list. The docs ask users to confirm in the bridge app that a specific pair has a route.

Weakest criteria: fees, audits and the pool route

Fee transparency at 2 is a deficiency. Synapse publishes no rate, no minimum and no split. Its own docs say only that the quote accounts for the bridge fee and that gas is collected when you sign. A reader cannot estimate the cost of a transfer before requesting a quote, which is the case the rubric scores lowest. Gas rebates and airdrops are mentioned, but not quantified.

Audit disclosure at 5 sits between disclosure and absence. The reports are real and name their firms, including a Trail of Bits comprehensive report and two Zellic assessments. They are filed in an assets folder of a developer monorepo, though, with no page in the docs pointing to them, no dates in most filenames and no mapping to the contracts that hold bridged funds. One ChainLight file is headed as a draft issue list. A reader who starts from the bridge cannot find this record.

Trust model disclosure at 6 is split by route. For the Intent Network, the docs set out what a Guard checks during the dispute period, what a dispute does, and that only Synapse runs Guards and Cancelers today. Unfilled deposits stay in escrow and can be cancelled permissionlessly after a longer delay. For the pool route, the Synapse Router page says only that assets are "moved to the destChain." It does not explain who attests the transfer or what secures nUSD and nETH.

What the score does not cover

• The brand change. synapseprotocol.com now redirects to hypercall.xyz/synapse, an options venue governed by the same DAO, while the bridge app, explorer and docs remain live at their own addresses. The rubric scores the bridge's documentation, not the brand's direction.
• Past incidents, which the category does not score unless the protocol documents them itself.
• Liquidity depth in the nUSD and nETH pools, and relayer competition on RFQ routes.
• Circle's own trust model on CCTP legs.

Our guide to how bridges work and why they are risky covers the pool and relayer failure modes in general terms.

Synapse vs Across and Squid

On current category marks, Across is sixth of nine at 7.2, Squid eighth at 6.0 and Synapse ninth at 5.8. Across runs the same optimistic relayer design as the Synapse Intent Network, and it scores higher on trust model (10 to 6), chain coverage (9 to 7), audits (8 to 5) and fees (5 to 2). Synapse leads it only on licence, MIT against Across's BUSL-1.1 (9 to 4). Squid splits the other way: it leads Synapse on fees (6 to 2), audits (8 to 5) and chain coverage (9 to 7), and trails it on trust model (5 to 6) and, badly, on licence (2 to 9).

Who this rating suits

The score is useful to a reader who values open code and wants to understand an optimistic relayer system with its single-operator caveat stated. It is weaker guidance for anyone who needs to know the cost of a transfer in advance or wants an audit record tied to the pool contracts. The rating is not a recommendation. On pool routes, check which intermediate token a transfer uses before signing, since sudden slippage can deliver nUSD or nETH instead of the token requested.