TGViewer
Channel Public Channel
Lighter API/MM Updates

Lighter API/MM Updates

@lighter_api_updates

Official Channel for API and Market-maker updates on Lighter
Subscribers
1.02K
Photos
1
Videos
1
Links
78

Showing posts older than #119 · Back to latest

Older Posts 20 shown
Post #118 785
Lighter API/MM Updates Today, we’re launching the Liquidity Partner Program, designed to reward participants who provide liquidity across RWA markets on Lighter. The program is open to anyone interested in supporting liquidity in these markets. We are focused on delivering best…
On a related note, Premium accounts will continue to enjoy 0 fees until Wednesday, April 15 at 14:30 UTC.
  • ❤ 2
Post #117 814
Today, we’re launching the Liquidity Partner Program, designed to reward participants who provide liquidity across RWA markets on Lighter. The program is open to anyone interested in supporting liquidity in these markets.

We are focused on delivering best-in-class liquidity to ensure traders can enter and exit positions in these markets efficiently.

This program is live; the full details are available in our public documentation: https://docs.lighter.xyz/liquidity-partner-program
docs.lighter.xyz Liquidity Partner Program | Lighter Docs
  • ❤ 4
Post #116 849
What is live after today's upgrade:

- You can now use Client Order IDs (CID) to group related orders (e.g., TP/SL). This lets you deterministically link them without relying on post-submission data.
- Maximum pending orders increased from 200 to 500.
- Other performance improvements.
  • ❤ 7
Post #115 795
You can now use the partnerStats endpoint to retrieve builder activity stats under the Partner Attribution program: https://apidocs.lighter.xyz/reference/partnerstats.

Additionally, builders managing multiple clients can authenticate each request to bypass IP-based rate limits, so that only L1-based limits apply. The builder tier is not required for this.
  • ❤ 5
Post #114 891
From now on, clients on a premium account will be able to mark some api keys as maker-only, restricting them to 0ms speed bump transactions only. Those can include:

- Post-only (ALO) orders
- Modify orders on ALO orders
- Cancel, and Cancel All orders

The direct benefit of using this configuration is having access to an optimized execution path, with reduced scheduling latency.

Implementation notes:
- Order types with any type of speed bump sent via a maker-only key will be rejected (error 62007)
- We apply a 1-hour cooldown in between api key type changes (error 62008)
- Maker-only api keys are only accessible by premium account types. Standard accounts will be rejected (error 62009)
- Api key indexes reserved by our front-end cannot be switched to maker-only (0-3)
- A maximum of 200 api keys can be set as maker-only, per account index
- It might take around a second for the api key type change to be effective
- Setting an array of keys as maker-only replaces the previous restrictions. Sending an empty array clears all maker-only restrictions
- To set an api key as maker-only, and get a list of them, we added POST /api/v1/setMakerOnlyApiKeys and GET /api/v1/getMakerOnlyApiKeys (both auth-gated): https://apidocs.lighter.xyz/reference/setmakeronlyapikeys ; https://apidocs.lighter.xyz/reference/getmakeronlyapikeys
  • ❤ 8
Post #113 964
- Lighter's GO & Python SDKs have been updated to support skipping nonces. Examples on how to skip nonces, and how to use grouped orders with partner codes have been added: https://github.com/elliottech/lighter-python/tree/main/examples

total_funding_paid_out has been added to the account endpoint, and the account_all WebSocket channel, allowing to fetch total funding paid out for an existing position without requiring authentication

- We're now able to share an S3 bucket containing all trade events going back to mainnet genesis with interested parties, updated daily. You can reach out via Discord #support to obtain access
  • ❤ 6
  • 👍 2
Post #112 988
Recent updates:

- Clients are now allowed to skip nonces. This is optional, and can be done by setting the SkipNonce field in L2TxAttributes to 1. We still require the next nonce to be higher than the previous, and lower than 2^47-1. SDKs support will be rolled out later this week
- The number of available sub-accounts has been increased from 8 to 64 on the highest staking tier, or fee credits equivalent
- The builder tier is now available, allowing 240K weighted requests on read endpoints, at no extra cost. You can reach out via Discord support channels to be part of the program
- The order_book and ticker WebSocket channels now include last_updated_at, a committer-level microseconds timestamp
- liquidation_price is now correctly displayed on WebSocket channels that return position data
  • ❤ 8
  • 👍 2
Post #110 1.05K
The improved WebSocket logic has been deployed to mainnet.

Find all details for the changes here: https://t.me/lighter_api_updates/101

Additionally, for a few days you will be able to fallback to the old ping/pong logic (other changes will still apply) adding this query parameter to the WebSocket URL: server_pings=true
  • 👍 6
Post #108 1.22K
- Volume quota cap has been increased from 5M to 15M
- This change: https://t.me/lighter_api_updates/101 to our WebSocket logic has been postponed to next Monday, March 23rd at 12pm UTC. Testnet already has the new logic, and you can fallback to the old (current) ping/pong logic adding server_pings=true to the WebSocket URL. This fallback will be made available on mainnet too, for a few days after we deploy the changes.

Other changes:
- You can now use the active_only filter on the account endpoint to hide markets for which leverage and margin settings are present (meaning, you traded the market at least once), but with no active position at the time of your request
- The logs endpoint for markets via the explorer api now supports pagination: https://apidocs.lighter.xyz/reference/get_markets-symbol-logs
- Maker trades are now available via the explorer api
  • ❤ 7
Post #107 1.08K
FYI: We will be updating our WTI methodology to a continuously rolling method in the near future. Starting Monday March 16th, WTI will track 100% CLK6 until further notice, expected in early April. We will keep you updated.
  • 👍 6
  • ❤ 1
  • 👎 1
Post #106 1.24K
Currently, only api_key_index:0 and api_key_index:1 are being reserved to serve desktop and mobile users on the lighter front-end

Starting March 22, 2026 at 2:30 PM UTC, the following indexes will be reserved to serve web and mobile users:

- api_key_index:0
- api_key_index:1
- api_key_index:2
- api_key_index:3

If you're currently using indexes 2, 3 for programmatic api trading, you should switch to any other other index in the 4-254 range to avoid any disruption.
  • 👍 4
Post #105 1K
Recent updates:

- Volume quota generation has been improved, now requiring $5 of notional traded to increase quota by one. Complete details are available here: https://apidocs.lighter.xyz/docs/volume-quota-program
- The accountActiveOrders endpoint does not require a market_id query parameter anymore, defaulting to all markets
- The markets/logs explorer endpoint now allows querying spot markets too, using the base-quote format, e.g. eth-usdc
  • ❤ 9
Post #103 885
The following WebSocket channels now send a millisecond timestamp attached to subscribe and update messages:
- ticker
- market_stats
- spot_market_stats
- account_all_assets
- account_spot_avg_entry_prices
- user_stats
- height

Recent updates:
- CCXT added Lighter support: https://t.me/ccxt_announcements/107
- Docs have been updated to include partner attribution integration details: https://apidocs.lighter.xyz/docs/partner-integration
- Python & Go SDKs have been updated to support partner attribution. Moreover, APIs have been updated and support for staking & unstaking has been added
  • ❤ 2
Post #102 786
Quick heads up — we’ll run a short maintenance at 12:59AM UTC (in 20mins). It should last around 2 minutes, after which everything will be back to normal. Thanks for your patience.
  • 👍 3
Post #101 2.79K
On March 20, 2026 at 9:30 PM UTC, the following WebSocket changes will go live to improve performance and stability:

- The server will no longer send application-level ping messages
- To keep the connection alive, clients will be expected to send a WebSocket ping frame (or any message) at least once every 2 minutes
- If the server does not receive any frames from a client for 2 minutes, the connection will be closed
- Clients that fall behind on reading messages (e.g. not consuming order book updates quickly enough) may be disconnected more aggressively than before
- The 24-hour forced connection lifetime will be removed, so well-behaved connections will be able to remain open indefinitely
- WebSocket compression (permessage-deflate) will be enabled, so clients that support it can benefit from reduced bandwidth usage

Additionally, the above changes will be deployed on testnet in a few days, allowing ample time to test the new logic.
  • 👍 5
  • ❤ 4
Post #100 845
Rate limits for sending transactions have been improved, providing up to an order of magnitude increase for the best staking tier. These changes apply to premium accounts only, and are tied to staked $LIT. Purchasing fee credits counts as staked $LIT for this purpose.

Changes include:
- Separate rate limit buckets. Other GET/POST endpoints and WebSocket channels will fall into a separate bucket, with unchanged rate limits. Getting rate limited in one of the two buckets will not affect activity on the other one
- sendTx and sendTxBatch will now only have L1 checks, while IP checks have been lifted

Complete details can be found in our api docs: https://apidocs.lighter.xyz/docs/rate-limits
  • ❤ 5
Post #99 786
Update re: Volume Quota

Going forward, only order-related transactions will count toward your volume quota consumption. All other transaction types will have separate rate limits and will no longer consume quota. The following transaction types will continue to count toward volume quota:

L2CreateOrder
L2CancelAllOrders
L2ModifyOrder
L2CreateGroupedOrders
  • 👍 4
Older posts →
Threads Profile ViewerView any public Threads profile without an account.Open ThreadLook →Writing with AI? Make it sound human.Metric37 rewrites AI drafts so they read naturally. Free AI detector, 1,500 words free.Try Metric37 →