x402engine retweeted
Most active x402 projects on Solana ⚡️
BlockRun | @BlockRunAI
Vishwa| @Vishwa_lab
thre | @trythreews
Xona Agent| @xona_agent
Agent402 | @Agent402Tools
SolEnrich | @SolEnrichHQ
Otto AI| @useOttoAI
AgentMail | @agentmail
Scry Wallet
CULT | @thecultos
x402engine | @x402engine
Exa| @ExaAILabs
fetcher
Bitrefill | @bitrefill
x402engine retweeted
remember you can make you agent autonomous via @x402engine paying with usdm on @megaeth
x402engine retweeted
working on it
@x402engine :)
x402engine retweeted
x402 Engine Goes Live with 38 Production APIs Behind HTTP 402 Micropayments
38 production APIs behind HTTP 402 micropayments - 15 LLMs, web search, scraping, TTS. Pay in USDC on Base, USDm on MegaETH, or USDC on Solana. No API keys required.
via @x402engine
x402engine.app/
x402engine retweeted
.@x402engine coded
a couple of random thoughts on agents:
today, specialized agents built for specific tasks will perform better than generalized agents, ie bankr agent is built for crypto and openclaw is not.
however this won't always be the case. the intelligence underlying general agents will get so good that a single agent will be able to do anything a specialized agent can do. they will create their tools on the fly to accomplish any task.
this also relates to agent coordination. what tasks can one agent do that another cannot? at some point a single agent will be able to do anything another agent can do without issue and coordination will not be required.
agents will use existing services, CLIs, and APIs available on the web. this is where x402 makes sense. x402 will be much faster for an agent to use vs communicating with another agent. free services via APIs and CLIs and paid services via x402 is what agents will use. the human to agent communication will be the driver not agent to agent.
this is why we are betting big on the self funding flywheel and cli tools we are building with bankr. they never become obsolete.
x402engine retweeted
.@x402engine now ships a signed agent manifest.
61 endpoints, each with: pricing, latency SLOs (p50/p95), determinism level, idempotency, fallback
equivalents, error taxonomy, refund policy, and network support.
SHA-256 hashed. ed25519 signed. machine-verifiable.
your agent can trust what it reads before it pays.
x402engine.app/.well-known/a…
x402engine retweeted
.@x402engine coded :)
let your agent autonomously pay for 60+ of the most popular endpoints in the internet
x402 is the endstate of agentic payments.
For a multitude of reasons.
Stripe calls them "machine payments," which is probably the best mental model for why they'll take over everything despite the cries for credit card sub accounts.
Here's a comparison of the 3 scenarios:
→ Credit Card Model (by human)
→ Credit Card Model (by agent)
→ x402 (by agent)
The reality is dealing with credit cards (subaccounts or not) and frontend UIs as a means to purchase something as an non-human is rife with legacy steps that are nonsensical.
They have to access the website, navigate the UI, input irrelevant form data (address? Name?) and deal with middlemen fees + settlement.
That all goes away with x402.
It goes away because it's a network-level communication and payment method which allows for the best of all UX:
→ It's cheaper (1-2% CC fee goes sub-%)
→ Settlement is instant
→ It requires no sign-up or login
When loyalty and switching costs goes to ~0 with agents they will choose the most pragmatic approach and it is VERY obvious it will be x402.
x402engine retweeted
turn your @openclaw into an autonomous agent
our @openclaw skill is live.
your agent asks a question → auto-discovers the right API → pays sub-cent via x402 → returns the
answer.
70+ APIs. invisible payments. no API keys to manage.
the agent sees answers, not blockers.
docs: x402engine.app/docs/skills/x…
x402engine retweeted
your agent can use exa without the need for an api key via @x402engine :)
x402engine retweeted
crypto price feeds on @x402engine cost $0.001. minimax 2.5 costs $0.01.
51 of the most popular endpoints on the web
pay with usdc on @base or @solana. or usdm on @megaeth.
x402engine retweeted
.@x402engine coded
x402engine retweeted
.@x402engine coded
x402engine retweeted
.@x402engine update: 28 llm models now live.
added 8 new models: gpt-5.2, gpt-5.2 codex, kimi k2.5, minimax m2.5, glm-5, grok code fast, seed 1.6, and devstral 2. all pay-per-call, same format.
also enabled ENS resolution, resolve any .eth name or reverse-lookup an address for $0.001.
47 apis total. no api keys, no accounts. live on @base @megaeth @solana
x402engine.app
x402engine retweeted
20 llms behind one payment gateway @x402engine. claude opus, gpt-5, o3, deepseek r1, gemini pro, perplexity with search. agents pick the model and pay per call. no api keys, just usdc. x402engine.app
x402engine retweeted
.@x402engine update: we have 20 llm models now
started with 5 this morning. got carried away.
claude, gpt, llama, mistral, gemini, deepseek, qwen. all pay-per-call with stablecoins.
agents have options now. on @megaeth @base @solana
x402engine retweeted
. @x402engine coded
Tempo's Dan Romero explains why the future of AI agents will be "stablecoin-native."
John: "Why wouldn't I just give OpenClaw my credit card?"
@DWR: "The credit card itself is kind of a private key, having that get prompt-injected out, maybe not the best thing in the world."
“If you have agent swarms, the idea of spinning up a new credit card for each individual sub-agent doesn’t make sense. With wallets, you can spin up as many as you want and manage the balances for each agent.”
“An API call to any of these frontier labs right now is pay-per-call. You’re eventually going to get to a point where, for every single API call, you can just pay some amount of stablecoins in the background and keep moving.”














