EcoCompute Should-I-Quantize Estimator
Energy impact of LLM quantization, calibrated to measured GPU data. Every answer states its basis.
Connect Remote server
claude mcp add --transport http ecocompute-estimator https://ecocompute-estimator.zhanghongping1982.workers.dev/mcp
{
"mcpServers": {
"ecocompute-estimator": {
"url": "https://ecocompute-estimator.zhanghongping1982.workers.dev/mcp"
}
}
}
{
"servers": {
"ecocompute-estimator": {
"type": "http",
"url": "https://ecocompute-estimator.zhanghongping1982.workers.dev/mcp"
}
}
}
Settings > Connectors > Add custom connector
URL: https://ecocompute-estimator.zhanghongping1982.workers.dev/mcp
Tools 4 found in our last check
estimate_quantization_energyEstimate the task-level energy change (ΔE% vs FP16) of weight-only quantization (NF4 or INT8) for a model of a given size on a GPU architecture class, from measured-anchored curves. The answer carries basis (measured/interpolated/estimated), confidence and evidence boundaries.
get_quantization_crossoverGet the model-size crossover for NF4/INT8 per architecture: below it quantization costs energy, above it saves. Crossovers belong to a card + software stack, not an architecture — the answer says so.
list_data_coverageList what is actually measured (GPU architecture × precision × size range, NVML direct), what is borrowed/estimated, and where the gaps are. Use this before trusting an estimate at the edges of coverage, and to answer 'is there data for my GPU?' honestly.
search_measurementsSearch the measured anchors by model name, GPU name, architecture or size. Returns ΔE% vs FP16 per anchor with run counts — the raw evidence behind the curves.
Details
| Endpoint | https://ecocompute-estimator.zhanghongping1982.workers.dev/mcp (Streamable HTTP) |
| First published | 24 Oct 2026 |
| Last updated | 24 Oct 2026 |
| Registry name | io.github.hongping-zh/ecocompute-estimator |
More ai and media servers See all
Listing data from the official MCP registry. Status and tools from NestDaddy's daily check. Browse all MCP servers.
