NoGPU
Server Details
Free report: how many of your LLM's classification calls a small CPU model can answer alone.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- Repository
- doronp/nogpu-mcp
- GitHub Stars
- 0
- Server Listing
- nogpu-mcp
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 4 tools
Each tool maps to a distinct step: nogpu_report creates a job, nogpu_status polls it, nogpu_buy fetches the purchase offer, and nogpu_delete removes data. The only overlap is that nogpu_status also returns report fields ('same fields as nogpu_report'), but the descriptions explicitly frame one as creation and the other as polling, so misselection risk is low.
All four tools follow a strict nogpu_<verb> pattern (buy, delete, report, status) with a shared server prefix. The convention is predictable and readable throughout.
Four tools is on the lean side but each covers a real lifecycle step (submit, poll, purchase, delete) with no filler. It could arguably use one more (e.g. listing or cancelling jobs), but the set is well-scoped for the service.
The surface covers the core flow: create a report job, poll status/retrieve results, get the paid offer with payment details, and delete data. Minor gaps exist (no job listing, no explicit cancel or re-run), but agents can work around these via job_id and claim_token.
Available Tools
4 toolsnogpu_buyBRead-onlyInspect
Read-only, never pays: for a certified (fit) report, returns the $99 offer to keep the model, the question to ask the user word for word, a card checkout link for the user, and the USDC (x402) payment details when the server offers them. If the job is not buyable, returns the reason.
Inputs: job_id (required), claim_token (required).
nogpu_deleteBDestructiveIdempotentInspect
Delete an uploaded job's data now (otherwise deleted within 24h, or 24h after payment if you bought the model).
Inputs: job_id (required), claim_token (required).
nogpu_reportAInspect
Free report: send an LLM's past decisions, 160-10,000 rows with at least 2 labels, either as rows ([{"text": ..., "label": ...}]) or as jsonl (the same rows as JSONL text; a JSON array or a CSV with a header row also works), at most 20 MB. NoGPU trains a small CPU model on them and measures what share of the decisions it answers alone, how often those answers match the LLM (with a 95% lower bound) on held-out rows, and its latency. Waits up to 40 seconds; if the report is not ready by then, call nogpu_status with the returned job_id and claim_token. claim_token and report_url (the token is after its #) open the job: treat both as secrets. Free reports through this endpoint come from one shared daily pool. Data is deleted within 24h. source says where the rows came from: agent (default), the app's stored history, rows logged by a collect snippet, or a given file.
Inputs: rows, jsonl, source.
nogpu_statusARead-onlyInspect
A job's status, and its report once ready (same fields as nogpu_report). Also shows download_url after payment: GET it with header X-Claim-Token. claim_token and report_url are secrets.
Inputs: job_id (required), claim_token (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Benchmark your own prompt on 100+ LLMs with real API calls; pick the cheapest model that passes.
Checks if a cheaper model agrees with your GPT, Claude or Gemini outputs, and what routing saves.
Sort up to 1,000 texts into your own labels with a calibrated confidence per answer. No API key.
Cost-optimized LLM model routing recommendations for autonomous AI agents
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables LLM agents to make fast semantic decisions by choosing one of 2–16 options with probabilities, via tools for single and batch classification.MIT
- AlicenseCqualityCmaintenanceAn MCP server that records the decision and categorisation steps inside agent workflows — inputs, choice and outcome — then serves confident decisions without an LLM call, routing uncertain ones to a hosted model and only training a private probe when reports show it's warranted. It reports agreement, coverage, savings, drift and hosted spend, and includes a drop-in gateway so pipelines making separate LLM classification calls can be answered without code changes.26MIT
- AlicenseAqualityBmaintenanceProvides fast, offline typed decision tools (classify, score, check, triage) to agents via the Model Context Protocol, enabling safe local inference across 100+ languages with calibrated confidence and no text generation.860 PyPIMIT
- AlicenseNot gradedqualityCmaintenanceEnables AI agents and IDEs to classify developer intent locally in under 2 milliseconds, executing tools instantly when confident and abstaining to a cloud LLM when uncertain. It also logs outcomes to a private on-device replay buffer for continuous fine-tuning and reports lifetime token, latency, and cost savings.1MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.