Aiprox MCP vs XFuel
A side-by-side look at the Aiprox MCP and XFuel MCP servers: transport, auth, tools, quality signals, and the config to install either one in Claude, Cursor, Windsurf or VS Code.
Verdict: choose Aiprox MCP for finance & fintech work as a local process; choose XFuel for finance & fintech as a local process. You can also run both side by side in one mcpServers block.
Which MCP server should you choose?
Choose Aiprox MCP when
- You need dedicated capabilities in the Finance & Fintech domain.
- You prefer a local stdio subprocess transport.
- Its auth model fits you: API key required.
- You want tools like
list_agents, get_agent, find_agent.
Choose XFuel when
- You need dedicated capabilities in the Finance & Fintech domain.
- You prefer a local stdio subprocess transport.
- You want tools like
chat_completions, submit_inference, register_agent.
Feature & specification comparison
Rows where the two servers differ are marked.
| Spec | ||
|---|---|---|
| Category | Finance & Fintech | Finance & Fintech |
| Quality signal | 56/100 · Good | 60/100 · Good |
| Install command | Confirmed | Confirmed |
| Transport | Local subprocess (stdio) | Local subprocess (stdio) |
| Auth | API key required | Not declared |
| Pricing | Paid service | Not declared |
| Required env vars | None listed | None listed |
| Verified owner | Community listing | Community listing |
| GitHub stars | 1 | 1 |
| Directory activity | 2 views 0 installs 0 upvotes | 2 views 0 installs 0 upvotes |
| Tools | 5 | 17 |
| Full listing | View Aiprox MCP | View XFuel |
Tools & capabilities
Aiprox MCP tools 5
list_agentsList all agents, optionally filter by capability or railget_agentGet full details for a specific agentfind_agentFind the best agent for a taskregister_agentRegister your agent in the registryget_specGet the agent manifest specification
XFuel tools 17
chat_completionsGenerate text via Chit402's OpenAI-compatible POST /v1/chat/completions. This is the first-hour / demo path. With the demo key it is unmetered (rate-limited). It is NOT submit_inference and does NOT spend USDC. Args: - messages (required): OpenAI chat messages, e.g. [{role:"user", content:"Say hello"}] - model (optional): catalog id from list_models (default xfuel/auto) - max_tokens, temperature (optional) Returns the completion plus top-level task_id and verify_url (also under xfuel). proof_outcome may still be pending — use verify_receipt for offline binding checks. Do not wait in this tool.submit_inferenceSubmit a PAID M2M inference task (POST /task-request). Default rail is USDC on Base. Without a payer this returns HTTP 402. This is NOT the unmetered demo path — use chat_completions to generate text with the demo key. Pass messages or input so a live provider is actually called. input_hash/memo alone do not send a prompt to the model. Args: - model (string): live catalog id from list_models - sender (string): 0x address that owns/pays for the task - amount (string): USDC 6 decimals; 2000 = $0.002; minimum 2000 - messages / input (optional): the prompt. Required for live routing. - max_tokens, temperature (optional) Poll get_task_status(task_id). Pay the live $0.01 door at POST /v1/chat/completions from the agent's wallet — MCP does not take a human private key.register_agentPOST /v1/agents/register. Binds an AAWP official or smart-account agentWallet to an integer agent_id using a collected HMAC-valid receipt (task_id from a paid POST /v1/chat/completions). Demo receipts do not qualify. API key is not a wallet. Does not accept a human private key. Returns agent_id (required later by POST /erc8004/validate) and validate_score.get_agent_bookGET|POST /v1/agents/:agent_id/book. Possession-gated last-N collected UsageSettled rows for that agent_id (task_id, payment.ref, rail, amount, collected_at, route). Demo / unmetered rows never appear. Default N=50, max 200. Totals: count, USDC sum, by rail. Response includes caps (budget Y, spent, remaining) when set. Not a public index. API key is not possession. Does not accept a human private key. Pass the session issued by register_agent (or an HMAC over agent_id + window using that session). Unauth or wrong proof returns 401/403 with an empty body.get_bookGET|POST /v1/agents/:agent_id/book. Possession-gated last-N collected UsageSettled rows for that agent_id (task_id, payment.ref, rail, amount, collected_at, route). Demo / unmetered rows never appear. Default N=50, max 200. Totals: count, USDC sum, by rail. Response includes caps (budget Y, spent, remaining) when set. Not a public index. API key is not possession. Does not accept a human private key. Pass the session issued by register_agent (or an HMAC over agent_id + window using that session). Unauth or wrong proof returns 401/403 with an empty body. Alias of get_agent_book for Eliza / Cursor clients.verify_receiptFetch GET /receipt/:task_id?format=json and verify payment binding locally via @xfuel/verify — no trust in HTML, no private keys. This is the default "prove it" path for agents holding hub, model, amount, verify_url. Args: - task_id (preferred) or verify_url — one required - fetch_jwks (default true): load /.well-known/jwks.json for issuer signature check - check_payer (default false): on-chain USDC payer match (Base or Solana RPC) - check_nullifier (default false): SP1 nullifier anchor read (optional Tier-2) Returns { verify_url, receipt, verification: { overall, binding, issuer_signature, … } }. overall is verified | partial | failed. Unmetered demo receipts may be partial (no binding).get_task_statusGet the current status of a Chit402 task, including proof outcome and fee breakdown. Args: - task_id (string): from chat_completions (xfuel.task_id) or submit_inference Returns JSON: { task_id, status, proof_outcome, verify_url, message_type, chain_id, gross_amount, fee_amount, net_amount, fee_bps, payment_rail, payment_ref, result, sp1_proof, created_at, updated_at }. 'verify_url' is a public, no-auth receipt page you can open or share to prove settlement. 'status' reaches a terminal value ('completed' | 'fee_collected' | 'failed'); 'proof_outcome' is one of 'pending' | 'valid' | 'regenerable' | 'invalid'.get_proofFetch the SP1 ZK settlement proof for a settled task. Args: - task_id (string): from chat_completions (xfuel.task_id) or submit_inference Returns JSON: { task_id, status, proof_outcome, verify_url, payment_binding, sp1_proof: { proof, publicInputs, nullifier, provingTimeMs }, fee }. 'verify_url' is a public, no-auth receipt page you can open or share to prove settlement. The proof attests settlement metadata + a commitment to the output hash (NOT inference correctness). To validate it, use verify_proof(task_id). Fails if the task has not settled yet — poll get_task_status until proof_outcome is 'valid'.verify_proofIndependently verify a settled task's proof client-side (the "prove it" flow in one call): 1. a proof is present and proof_outcome === 'valid'; 2. the x402 payment binding (if present) re-derives to the committed value; 3. (optional) the on-chain nullifier's spent state, when check_nullifier=true AND the server was started with XFUEL_RPC_URL + ZK_VERIFIER_ADDRESS. Args: - task_id (string): from chat_completions (xfuel.task_id) or submit_inference - check_nullifier (boolean, optional): also read the on-chain nullifier state (default false) Returns JSON: { ok, checks: { hasProof, proofOutcomeValid, paymentBinding, nullifier }, reasons }. 'ok' is true when the proof is present + valid and any present payment binding is consistent. The nullifier read is informational and does not gate 'ok'.quote_taskPreview per-rail pricing WITHOUT creating a task. Use before submit_inference. Unknown model ids still return the $0.01 USDC floor — the human summary warns when priced_model is null (inference will fail). Args: - model (preferred) or model_id: catalog id to price - amount (optional): USDC 6 decimals for the USDC rail - messages / max_tokens / tools: forecast the request you will submit Unmetered chat_completions does not need this quote.get_healthCheck the connected Chit402 API's health and configuration (fee config, supported chains, message types, demo limits). Useful for discovery/diagnostics. Args: none. Returns JSON: the /health payload (status, server, version, fee_config, chains, message_types, …).get_my_statsFetch usage for the configured API key (GET /stats/me). The public demo key chit402-demo is SHARED — numbers are not this client. Bring your own XFUEL_API_KEY for buyer-scoped stats.- +5 more on the listing page
Ready-to-paste client configs
Paste either block (or both) into your client config, e.g. claude_desktop_config.json or ~/.cursor/mcp.json. We only show a config when the install command is confirmed.
Aiprox MCP
{
"mcpServers": {
"unixlamadev-spec-aiprox-mcp": {
"command": "npx",
"args": [
"-y",
"aiprox-mcp"
]
}
}
}
XFuel
{
"mcpServers": {
"xfuel": {
"command": "npx",
"args": [
"-y",
"chit402-mcp"
]
}
}
}
Frequently asked questions
Aiprox MCP is categorized under Finance & Fintech and uses a local stdio subprocess. XFuel belongs to Finance & Fintech and uses a local stdio subprocess. Pick Aiprox MCP for finance & fintech work and XFuel for finance & fintech.
Copy the mcpServers block from the config section on this page into your client's config (for example claude_desktop_config.json or ~/.cursor/mcp.json), then fully restart the client.
Aiprox MCP is listed as a paid service. Auth: api key required. XFuel has no declared pricing — check the project's own site. If a server needs environment variables such as API keys, define them under the env key in your client's MCP config.
Yes. MCP clients support multiple servers: add unixlamadev-spec-aiprox-mcp and xfuel as separate keys in the same mcpServers object, and your assistant will call tools from either one.
On AllMCPs, Aiprox MCP has a quality signal of 56/100 (Good) with 2 views, 0 installs and 1 GitHub stars. XFuel has 60/100 (Good) with 2 views, 0 installs and 1 GitHub stars.