About IBM Granite via x402
An MCP server that exposes IBM Granite models as pay-per-use tools through the x402 payment protocol. Callers pay in USDC on Solana and need no IBM Cloud account, while the operator supplies the watsonx.ai credentials and a receiving wallet. It suits developers who want to offer or use Granite inference on demand.
Key features
- 01Per-call USDC payment for Granite tools
- 02Chat, code, embeddings, document analysis and forecasting
- 03Free getting-started tool that explains prices and flow
- 04Tool annotations describing read-only and open-world behavior
- 05Automatic payment loop in x402-capable clients
Use cases
- 01Calling Granite models from Claude Code or Cursor on demand
- 02Producing structured analysis of documents with entities and risk flags
- 03Zero-shot forecasting of time-series data
- 04Generating batch text embeddings for RAG or search
MCP support
The server runs over stdio and exposes six tools to MCP clients. Unpaid calls return a 402 PaymentRequired envelope, and paid results include a settlement receipt in the response metadata.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y @three-ws/ibm-x402-mcpMCP_SVM_PAYMENT_ADDRESSSolana wallet address that receives USDC payments from tool callers (server operator's receiving wallet). Required to accept x402 payments.WATSONX_API_KEYSecretIBM Cloud API key used by the server to call watsonx.ai on behalf of end users. Create one at https://cloud.ibm.com/iam/apikeys.WATSONX_PROJECT_IDwatsonx.ai project id scoping all inference calls. Find it under your project → Manage → General → Project ID. Provide this or WATSONX_SPACE_ID.WATSONX_SPACE_IDwatsonx.ai deployment space id. Alternative to WATSONX_PROJECT_ID.WATSONX_URLRegional watsonx.ai inference host. Defaults to https://us-south.ml.cloud.ibm.com. Other regions: eu-de, eu-gb, jp-tok, au-syd, ca-tor.WATSONX_MODEL_IDDefault chat/generation Granite model id. Defaults to ibm/granite-3-8b-instruct.WATSONX_EMBED_MODEL_IDDefault embedding model id. Defaults to ibm/granite-embedding-278m-multilingual.X402_FEE_PAYER_SOLANASolana account that sponsors transaction fees for x402 payments. Defaults to the three.ws fee payer.X402_FACILITATOR_URLx402 facilitator endpoint for payment verification and settlement. Defaults to https://facilitator.payai.network.Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project README@three-ws/ibm-x402-mcp
Pay-per-use IBM Granite AI over MCP — chat, code, embeddings, analysis, and forecasting, billed in USDC on Solana. No IBM account required.
Install · Quick start · Tools · Payment flow · Requirements · three.ws
A Model Context Protocol server that exposes IBM Granite foundation models as pay-per-use tools via the x402 payment protocol. End users pay USDC on Solana per call — no IBM Cloud account of their own. The server operator supplies IBM credentials (
WATSONX_*) and a receiving Solana wallet (MCP_SVM_PAYMENT_ADDRESS); callers supply only USDC. One free tool (ibm_granite_getting_started) explains prices and the flow before any payment.
Built by three.ws. Community-built and not affiliated with IBM.
How it works
- An MCP client (Claude Desktop, Claude Code, Cursor, or an agent) connects to this server.
- The client calls a tool — e.g.
ibm_granite_chat. - Without an x402 payment payload, the server returns a
402 PaymentRequiredenvelope quoting the USDC price and the Solana receiving address. - The client signs a Solana USDC transfer and retries with the payment in
_meta["x402/payment"]. - The server verifies and settles the payment via the facilitator, calls IBM watsonx.ai, and returns the result with a settlement receipt in
_meta["x402/payment-response"].
x402-capable MCP clients handle this loop automatically.
Install
npm install @three-ws/ibm-x402-mcp
Run it directly with npx (no install needed):
MCP_SVM_PAYMENT_ADDRESS=<your-solana-wallet> \
WATSONX_API_KEY=<ibm-api-key> \
WATSONX_PROJECT_ID=<watsonx-project-id> \
npx @three-ws/ibm-x402-mcp
Or install globally for the ibm-x402-mcp binary on your PATH:
npm install -g @three-ws/ibm-x402-mcp
Quick start
Whoever runs the process is the operator: the server fails fast at startup without a payment address and IBM credentials (WATSONX_API_KEY, WATSONX_PROJECT_ID). Callers of the running server need only USDC — no IBM account. With Claude Code, one command:
claude mcp add ibm-granite-x402 \
-e MCP_SVM_PAYMENT_ADDRESS=your-solana-wallet \
-e WATSONX_API_KEY=your-ibm-api-key \
-e WATSONX_PROJECT_ID=your-watsonx-project-id \
-- npx -y @three-ws/ibm-x402-mcp
Or wire the server into your MCP client config (claude_desktop_config.json, Cursor's mcp.json):
{
"mcpServers": {
"ibm-x402": {
"command": "npx",
"args": ["-y", "@three-ws/ibm-x402-mcp"],
"env": {
"MCP_SVM_PAYMENT_ADDRESS": "your-solana-wallet-address",
"WATSONX_API_KEY": "your-ibm-cloud-api-key",
"WATSONX_PROJECT_ID": "your-watsonx-project-id"
}
}
}
}
Inspect the tool surface with the MCP Inspector:
npx -y @modelcontextprotocol/inspector npx @three-ws/ibm-x402-mcp
Tools
| Tool | What it does | Price |
|---|---|---|
ibm_granite_getting_started | Overview, prices, and the x402 payment flow. No payment or IBM account required. | Free |
ibm_granite_chat | Conversational AI via IBM Granite (default ibm/granite-3-8b-instruct). | $0.02 USDC |
ibm_granite_code | Code generate, review, refactor, explain, test, document. | $0.025 USDC |
ibm_granite_embed | Batch text embeddings for RAG, search, and clustering (1–64 texts). | $0.005 USDC |
ibm_granite_analyze | Structured document analysis: entities, sentiment, risk flags, summary, next steps. | $0.04 USDC |
ibm_granite_forecast | Zero-shot time-series forecasting via IBM Granite TTM (Tiny Time Mixer). | $0.05 USDC |
Every tool is a read-only model-inference call — nothing on your machine or in any account is modified — and declares MCP tool annotations (readOnlyHint, openWorldHint, idempotentHint) so clients can reason about side effects before paying.
Input parameters
ibm_granite_getting_started — section (optional: overview (default, everything) / pricing / payment / tools / setup).
ibm_granite_chat — messages (required: 1–50 { role, content } pairs; roles system/user/assistant, content ≤ 32,000 chars), model (default ibm/granite-3-8b-instruct), max_new_tokens (1–4096, default 1024), temperature (0–2, default 0.7).
ibm_granite_code — task (required: generate/review/refactor/explain/test/document), prompt (required, ≤ 16,000 chars: the description for generate, the code for everything else), language, context (≤ 4,000 chars).
ibm_granite_embed — inputs (required: 1–64 texts, ≤8000 chars each), model.
ibm_granite_analyze — document (required), analysis_type (general/contract/financial/technical/medical/sentiment, default general), language.
ibm_granite_forecast — timestamps (required: 64–1024 ISO-8601, uniform cadence, oldest first), values (required: 64–1024 numbers, same length), freq (required: pandas cadence, e.g. 1h, 1D), prediction_length (1–96), label.
Example calls
// ibm_granite_chat
{
"messages": [
{ "role": "system", "content": "You are an expert data engineer." },
{ "role": "user", "content": "Design a lakehouse schema for IoT sensor telemetry." }
],
"max_new_tokens": 1024,
"temperature": 0.7
}
// ibm_granite_code
{ "task": "review", "prompt": "def calculate_roi(revenue, cost): return revenue / cost", "language": "Python" }
// ibm_granite_embed
{ "inputs": ["enterprise data governance", "cloud-native AI pipeline", "real-time analytics"] }
// ibm_granite_analyze
{ "document": "This Software License Agreement is entered into between...", "analysis_type": "contract" }
// ibm_granite_forecast (timestamps/values must be 64–1024 points; abbreviated here)
{ "timestamps": ["2025-01-01T00:00:00Z", "...", "2025-03-05T00:00:00Z"], "values": [12500, "...", 13200], "freq": "1D", "prediction_length": 14, "label": "daily_revenue_usd" }
Payment flow
This server uses the x402 protocol for micropayments:
- Client calls a tool without payment →
402 PaymentRequiredwith the USDC amount and Solana address. - Client builds and signs a Solana USDC transfer transaction.
- Client retries with the signed tx in
_meta["x402/payment"]. - Server verifies and settles via the configured facilitator (default PayAI).
- Server calls IBM watsonx.ai and returns the result with
_meta["x402/payment-response"](settlement receipt).
MCP Client (Claude Desktop / Cursor / agent)
│ tools/call (with x402 payment in _meta)
▼
ibm-x402-mcp (stdio MCP server)
│ verify + settle USDC on Solana
├──► x402 facilitator (default https://facilitator.payai.network)
│
│ inference call with IAM Bearer token
└──► IBM watsonx.ai (us-south.ml.cloud.ibm.com)
└── IBM Granite 3 8B Instruct / Embedding / TTM
Requirements
- Node.js >= 20.
- A Solana wallet address to receive USDC (
MCP_SVM_PAYMENT_ADDRESS). - IBM Cloud credentials: an API key (create one) and a watsonx.ai project id (Project → Manage → General → Project ID), or a deployment space id.
Environment variables
| Variable | Required | Default |
|---|---|---|
MCP_SVM_PAYMENT_ADDRESS | yes | — (aliases: X402_PAY_TO_SOLANA, X402_PAY_TO) |
WATSONX_API_KEY | yes | — |
WATSONX_PROJECT_ID | yes (or WATSONX_SPACE_ID) | — |
WATSONX_SPACE_ID | alternative to WATSONX_PROJECT_ID | — |
WATSONX_URL | no | https://us-south.ml.cloud.ibm.com |
WATSONX_MODEL_ID | no | ibm/granite-3-8b-instruct |
WATSONX_CODE_MODEL_ID | no | ibm/granite-3-8b-instruct |
WATSONX_EMBED_MODEL_ID | no | ibm/granite-embedding-278m-multilingual |
WATSONX_FORECAST_MODEL | no | ibm/granite-ttm-512-96-r2 |
WATSONX_TIMEOUT_MS | no | 90000 |
X402_FEE_PAYER_SOLANA | no | three.ws fee payer |
X402_FACILITATOR_URL | no | https://facilitator.payai.network |
X402_FACILITATOR_TOKEN | no | — (Bearer token for a private facilitator) |
X402_ASSET_MINT_SOLANA | no | canonical Solana USDC mint |
Regional hosts: us-south, eu-de, eu-gb, jp-tok, au-syd, ca-tor — e.g. https://eu-de.ml.cloud.ibm.com.
Errors
Every failure state is explicit — there is no silent fallback:
| State | What you see | Recovery |
|---|---|---|
| Missing env at startup | The process exits with a one-line configuration error: … naming the missing variable (MCP_SVM_PAYMENT_ADDRESS, WATSONX_API_KEY, or WATSONX_PROJECT_ID/WATSONX_SPACE_ID). | Set the variable and restart. |
| Unpaid call to a paid tool | An x402 PaymentRequired envelope quoting the USDC price and pay-to address — not an error; it is step 1 of the payment loop. | Pay and retry (x402-capable clients do this automatically). |
| Invalid/underpaid payment | The facilitator rejects verification and a fresh PaymentRequired envelope is returned; no inference runs and no funds settle. | Re-sign against the quoted accepts. |
| IBM-side failure | Tool result { "ok": false, "error": "watsonx_error", "message": "…", "status": … } — IAM auth failure (bad WATSONX_API_KEY), model error, or a watsonx.ai timeout (WATSONX_TIMEOUT_MS, default 90 s). | The message carries IBM's reason; check credentials, project id, and region. |
| Unexpected server fault | Tool result { "ok": false, "error": "internal_error", "message": "…" }. | Retry; file an issue if it persists. |
| Facilitator unreachable at boot | A facilitator init warning: … line on stderr; the server still starts and retries lazily on the first paid call. | Check X402_FACILITATOR_URL connectivity. |
ibm_granite_getting_started is always free and never touches IBM or the facilitator — use it to sanity-check the server before paying.
Related
@three-ws/ibm-watsonx-mcp— the same IBM Granite tools driven by your own IBM Cloud credentials (no x402, no per-call payment).
Links
- Homepage: https://three.ws
- Changelog: https://three.ws/changelog
- Issues: https://github.com/nirholas/three.ws/issues
- License: Apache-2.0, see LICENSE
Part of the three.ws SDK suite — 3D AI agents, on-chain identity, and agent payments.
Website · Changelog · GitHub
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.