Every change Wellknown observed on this MCP server, newest first, with what it was before and what it became. Tool-surface changes carry the definition diff. Nothing here is edited after the fact.
Added "estimate_gpu_cost", "estimate_token_cost", "fetch" and 16 more; changed the definition of "get_gpu_prices", "get_price_history" and "get_token_prices" (3 tools before, 22 now)
LatestToday's GPU rental prices: the price index in USD per GPU-hour for each GPU (H100, H200, A100, B200, B300, RTX PRO 6000), with1D1-day,1W1-week and1M1-month changes, the index day and a link to each GPU's page. Give one GPU to also get its benchmark contract, each provider'sstandardizedaverage price standardized to that contract, and its compute futures prices: each exchange listing a contract on its rental price, with theexchange'slast price and 24-hour volume, readfrom its public market-data APIwithin the last minute (all of them: https://computescreener.com/futures). For past values use get_price_history; for providers' posted prices get_provider_prices; for a cost estimate_gpu_cost.
before
—after
{"properties":{"gpus":{"items":{"properties":{"benchmark_contract":{"type":"string"},"change_1d_pct":{"description":"Percent change from the value 1 day earlier","type":["number","null"]},"change_1m_pct":{"description":"Percent change from the value 30 days earlier","type":["number","null"]},"change_1w_pct":{"description":"Percent change from the value 7 days earlier","type":["number","null"]},"date":{"description":"The index day (US Central)","type":"string"},"gpu":{"type":"string"},"markets":{"items":{"properties":{"as_of":{"type":["string","null"]},"contract":{"type":"string"},"note":{"type":["string","null"]},"price":{"type":["number","null"]},"status":{"type":"string"},"symbol":{"description":"The exchange's ticker","type":"string"},"venue":{"type":"string"},"volume_24h_usd":{"type":["number","null"]}},"required":["venue","contract"],"type":"object"},"type":"array"},"note":{"type":"string"},"price":{"description":"Latest index value, USD per GPU-hour","type":["number","null"]},"provider_averages":{"items":{"properties":{"average_price":{"type":["number","null"]},"carried_from":{"type":"string"},"price_count":{"type":["number","null"]},"provider":{"type":"string"},"read_at":{"description":"When its prices were last read","type":"string"},"source_url":{"description":"Where its prices were read","type":"string"},"status":{"type":"string"}},"required":["provider"],"type":"object"},"type":"array"},"providers":{"description":"Providers in the index","type":"integer"},"provisional":{"description":"True while the day is not over","type":"boolean"},"url":{"description":"The GPU's page","type":"string"}},"required":["gpu","price"],"type":"object"},"type":"array"},"license":{"type":"string"},"methodology":{"type":"string"},"sources":{"type":"string"},"unit":{"type":"string"},"updated":{"description":"When the GPU prices were last calculated (ISO 8601, UTC)","type":"string"}},"required":["unit","updated","gpus"],"type":"object"}before
—after
{"properties":{"kind":{"enum":["gpu","company"],"type":"string"},"license":{"type":"string"},"name":{"type":"string"},"points":{"items":{"properties":{"date":{"type":"string"},"price":{"type":"number"},"provisional":{"type":"boolean"}},"required":["date","price"],"type":"object"},"type":"array"},"sources":{"type":"string"},"unit":{"type":"string"},"url":{"type":"string"}},"required":["name","kind","unit","points"],"type":"object"}LatestToday's AI tokenpricepricesindexbypercompany:AIeachcompanycompany's price index in USD per 1M tokens (3 input : 1 output blend, weighted by its models' usage), with1D1-day,1W1-week and1M1-month changes. Give one company to also get its models' usage shares and posted input and output prices. For one model's prices use get_model_prices; for past values get_price_history.
Estimate a GPU rental cost — What renting a GPU would cost for a number of hours: at the index price and at each provider's standardized average, cheapest first (on-demand, benchmark contract; a provider's bill depends on its own terms).
Estimate an AI API cost — What a number of input and output tokens would cost on each model at its posted prices, cheapest first, and on each company's models in their usage shares. Give millions of tokens (e.g. 30 input, 10 output).
Read a Compute Screener page — The full text of a result from search, by its id (gpu:H100, company:OpenAI, model:<page>, provider:<page>, page:methodology, page:futures, page:verification): prices, changes, history, providers or models, as on the site.
Which GPU fits a model — Which GPUs can hold a model of a given size and what that setup costs per hour: the memory needed (weights at the precision, plus room for the KV cache and activations, or optimizer state for training), how many of each GPU, whether they fit in one 8-GPU node, and the cost at the index and at the cheapest provider.
Data freshness and sources — How fresh the data is: when the GPU and token indices were last calculated, today's index day, each GPU price source's status and last read, each GPU index's provider count and imputed share, and the latest days with a manifest.
Compute futures curve — Compute futures on one GPU's rental price against its index: every exchange's contract (perpetual or dated), last price, 24-hour volume and premium or discount to the index, and the volume-weighted futures price. Read from the exchanges within the last minute.
Prices of every other GPU — Rental prices of GPUs beyond the index (RTX 5090, RTX 4090, L40S, RTX 6000 Ada, RTX A6000, A40, L4, GH200, V100 and about 30 more): each model's lowest, median and highest price per GPU-hour today as posted, its offers, how many are in stock, its providers and the cheapest, from RunPod, Vast.ai, Akash, Shadeform's catalog of clouds and Lium. Give a GPU for its daily history. Not an index: configurations are not standardized (the seven indexed GPUs are in get_gpu_prices).
GPU price at a time — A GPU's index price at a past date or time: the value calculated at or just before an ISO time within the last few days (prices are recalculated every minute; kept at least every 30 minutes), else the daily value of that US Central day.
GPU supply and volume — Supply behind each GPU index: offers read at the latest collection and how many had capacity to rent (in stock), prices used, providers read today or carried, prices left out by reason, and each exchange's 24-hour compute futures volume.
Market overview — The compute market in one answer: how many GPU and AI company indices rose or fell today, their average 1-day, 1-week and 1-month changes, the cheapest and most expensive, the largest moves, GPU offers in stock, compute futures volume and the day's AI token volume.
Price ranges and volatility — For each GPU and AI company index (or one): its 7- and 30-day averages, its 30- and 90-day high and low, its 90-day change, its 30- and 90-day volatility (standard deviation of daily changes, annualized), its largest daily move in 90 days, and how long it has been unchanged.
AI model API prices — Each AI model's posted API prices in USD per 1M input and output tokens (the models behind the token index), with its usage share, its last price change and its 30-day change. Give a model (name or page id, e.g. "Claude Opus 5.5" or "kimi-k3") for its daily price history and every price change, or a company to list its models.
What changed — What moved recently: every listed GPU and company index's change over the last few days, largest first, and every AI model whose posted prices changed in that time.
GPU prices by provider — Each GPU cloud's and marketplace's posted on-demand prices per GPU (lowest and highest that day, as posted, with the configuration of the lowest) and its average in the index, standardized to the GPU's benchmark contract so providers compare like for like. Give a provider (e.g. "CoreWeave", "RunPod") and/or a GPU.
AI token volume — How many tokens each AI company's models processed per day on OpenRouter (the volumes that weight the token index), with 1-week and 1-month changes, each company's share of the total, its daily series and, for the latest day, each model's tokens. Daily from 2025-01-01.
Inference economics per GPU — What serving AI tokens on each GPU costs and earns at today's prices: the break-even throughput (tokens per second one GPU must sell at a model's or company's posted price to cover its rental) and, given a throughput, the GPU's cost per 1M tokens, revenue per GPU-hour and gross margin.
Every GPU and company covered — Every GPU rental index and AI company token index Compute Screener covers: whether it is listed (has a price and a month of history), its latest value and day, when its history starts, its provider or model count, its benchmark contract (GPUs) and its page. Use it to see what the other tools accept.
Search Compute Screener — Search Compute Screener's GPUs, AI companies, AI models, GPU providers and methodology by keyword (e.g. "H100", "Claude Opus", "CoreWeave", "benchmark contract"). Returns ids for fetch.
Check a published price — How a published daily value can be checked: whether it is final, and for a final day its manifest (every file it was computed from with its SHA-256, the factors in force, the calculation code's SHA-256 and the final values) and the command that recomputes it from the public files with the published code. Omit date for the latest day with a manifest.
Certificate recorded, valid to 2027-01-12
Authorization not required
Changed the definition of "get_gpu_prices"
⟨30 unchanged words⟩ contract, each provider's standardized average price, andtheitsexchangescompute futures prices: each exchange listing a contract on its rental price:,eachwith the exchange's last price and 24-hour volume, read from its public market-data API within the lastminute.minute (all of them: https://computescreener.com/futures).
Unknown → Live
First tool surface recorded: 3 tools (server version 1.4.2)
https://computescreener.com/mcp (mcp_streamable_http) — from mcp_registry, with the record
Showing the latest 7 events. The API returns up to 500 and filters by kind: ?kind=tool_surface_changed
before
—after
{"properties":{"companies":{"items":{"properties":{"change_1d_pct":{"description":"Percent change from the value 1 day earlier","type":["number","null"]},"change_1m_pct":{"description":"Percent change from the value 30 days earlier","type":["number","null"]},"change_1w_pct":{"description":"Percent change from the value 7 days earlier","type":["number","null"]},"company":{"type":"string"},"date":{"description":"The index day (UTC)","type":"string"},"models":{"items":{"properties":{"blended_price":{"type":["number","null"]},"input_price":{"type":["number","null"]},"model":{"type":"string"},"output_price":{"type":["number","null"]},"usage_share":{"description":"Share of the company's tokens that day (0 to 1)","type":["number","null"]}},"required":["model"],"type":"object"},"type":"array"},"note":{"type":"string"},"price":{"description":"Latest index value, USD per 1M tokens (3 input : 1 output)","type":["number","null"]},"url":{"description":"The company's page","type":"string"}},"required":["company","price"],"type":"object"},"type":"array"},"license":{"type":"string"},"methodology":{"type":"string"},"sources":{"type":"string"},"unit":{"type":"string"},"updated":{"description":"When the token prices were last calculated (ISO 8601, UTC)","type":"string"}},"required":["unit","updated","companies"],"type":"object"}