Lantern Works
Free, keyless, read-only GPU rental price comparison checked daily, with cited rates and job-cost estimates. Install: uvx lantern-works-deals-mcp · Docs
Live Answered an MCP handshake without credentials and listed 13 tools.
- Category
- Cloud, ops & data
- Transport
http- Source
- https://github.com/connormatthewdouglas/agent-work-stack-deals
- Endpoint
https://lantern-works-mcp.lantern-works.workers.dev/mcp- Reports itself as
- lantern-works-deals 0.6.6
- Protocol
- 2025-06-18
- Tools
- 13
- Context cost
- about 4,740 tokens for the tool definitions (estimated at four characters per token)
- Tool definitions last changed
- 2026-11-20
- Handshake time
- 1017 ms
- Repository
- 2 stars · last push 2026-11-22 · MIT
- Last checked
- 2026-11-22 11:27:39 UTC
- Listed since
- 2026-11-18
Health and compatibility
| Endpoint reachable | Verified | Responded in 1017 ms |
|---|---|---|
| MCP handshake (initialize) | Verified | Completed without credentials |
| Tool discovery (tools/list) | Verified | 13 tools listed |
| Authentication | Verified | None needed to connect |
| Protocol version | Verified | 2025-06-18 |
| Transport | Verified | Streamable HTTP |
| Capabilities declared | Verified | tools |
Last checked . Verified means the check saw it happen; it says nothing about whether the server is safe.
Recent history
5 daily checks so far; median response 1017 ms. Green answered, violet asked for auth, red did not answer.
Set up in your client
Claude Code
Run this in a terminal:
claude mcp add --transport http lantern-works https://lantern-works-mcp.lantern-works.workers.dev/mcpClaude Desktop
Open Settings → Connectors → Add custom connector and paste this URL:
https://lantern-works-mcp.lantern-works.workers.dev/mcpOr bridge it through mcp-remote in claude_desktop_config.json (needs Node.js):
{
"mcpServers": {
"lantern-works": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://lantern-works-mcp.lantern-works.workers.dev/mcp"
]
}
}
}Cursor
Or add it to ~/.cursor/mcp.json (or .cursor/mcp.json in a project):
{
"mcpServers": {
"lantern-works": {
"url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
}
}
}VS Code
Or add it to .vscode/mcp.json:
{
"servers": {
"lantern-works": {
"type": "http",
"url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
}
}
}Or from a terminal:
code --add-mcp '{"name": "lantern-works", "type": "http", "url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"}'Windsurf
Add it to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"lantern-works": {
"serverUrl": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
}
}
}Other clients
Most clients accept this shape. The endpoint is:
https://lantern-works-mcp.lantern-works.workers.dev/mcp{
"mcpServers": {
"lantern-works": {
"type": "http",
"url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
}
}
}If it does not connect
- Restart the client after editing a config file, and keep keys out of files you commit.
- If the tools do not appear, open the client's MCP log: a 401 means the credentials were not accepted, and a timeout means the server did not answer.
Tools 13
Names, descriptions and input schemas exactly as the server returned them on 2026-11-22. This is the server author's own text. Nothing is run from this page. Labels such as read-only are declared by the server and are not verified. Raw JSON
compare_dealsread-onlyexternal accessCompare 2-5 deals side by side (price, unit, offer type, code, verification), best value first.
Compare 2-5 deals side by side (price, unit, offer type, code, verification), best value first.
| Parameter | Type | Description | |
|---|---|---|---|
ids | string | required | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "compare_deals",
"arguments": {
"ids": "..."
}
}
}compare_gpu_pricesread-onlyexternal accessCompare tracked GPU rental prices for one model (e.g. 'H100', 'A100 SXM', 'RTX 4090', 'L40S'), cheapest …
Compare tracked GPU rental prices for one model (e.g. 'H100', 'A100 SXM', 'RTX 4090', 'L40S'), cheapest first. If a result looks wrong, give the user its report_issue_url. Same data as https://thelanternworks.com/api/v1/gpu-prices.json. Sorted by hourly_usd ascending; each result has source_url and checked_at. 'Cheapest' means cheapest among tracked providers, not the whole market. limit 1-25. Unknown model returns not_found listing tracked models. H100/A100: cheapest_by_form_factor gives one pick per form_factor (tier_class, interruptible, in_stock); same rule as the site. 'NVLink' rows are form 'not stated': NVLink is the bridge, and the vendor doesn't say SXM or PCIe. Each result has tier_class: on-demand | community | marketplace | interruptible (spot/preemptible) | reserved | dedicated-monthly. Reserved (reservation/commitment 'starting from') rates are left out unless include_reserved=true, and even then they are never reported as cheapest. DEFAULT on_demand_only=true: results[0] is the cheapest on-demand rate. Pass on_demand_only=false to include spot/interruptible, community and marketplace capacity (each such row is labelled by tier_class and interruptible); the summary then names the cheapest on-demand rate when the overall cheapest isn't. If a model has no on-demand rate, or every on-demand rate needs more GPUs than requested, the default falls back to all rates and says so in on_demand_fallback. gpus (optional, 1-1024; default 1, stated in assumptions): the GPU …
| Parameter | Type | Description | |
|---|---|---|---|
gpu | string | required | |
limit | integer | optional | |
include_history | boolean | optional | |
on_demand_only | boolean | optional | |
include_reserved | boolean | optional | |
max_data_age_hours | number | optional | |
region | string | optional | |
min_usable_memory_gb | number | optional | |
gpus | integer | optional | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "compare_gpu_prices",
"arguments": {
"gpu": "..."
}
}
}create_watch_keycan modifyexternal accessHosted only, free, opt-in. Register a webhook for GPU price alerts and get an api_key plus a webhook signing …
Hosted only, free, opt-in. Register a webhook for GPU price alerts and get an api_key plus a webhook signing secret (both shown ONCE; store them). webhook_url must be public https. email is optional contact only (no mail is sent). No payment. Stored: a hash of the key, the webhook URL, optional email, your watches and per-key counts.
| Parameter | Type | Description | |
|---|---|---|---|
webhook_url | string | required | |
email | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "create_watch_key",
"arguments": {
"webhook_url": "..."
}
}
}delete_watchcan modifyexternal accessHosted only. Delete one of your watches by watch_id.
Hosted only. Delete one of your watches by watch_id.
| Parameter | Type | Description | |
|---|---|---|---|
watch_id | string | required | |
api_key | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "delete_watch",
"arguments": {
"watch_id": "..."
}
}
}estimate_job_costread-onlyexternal accessCost of one GPU job (not cost per hour) among tracked providers, cheapest current compute subtotal first. …
Cost of one GPU job (not cost per hour) among tracked providers, cheapest current compute subtotal first. gpu_family (e.g. 'H100', 'A100 80GB') or its alias gpu (same name compare_gpu_prices uses) is required; if both are given they must match, else invalid_argument. gpus_per_job defaults to 1. Ranking matches compare_gpu_prices' default: fresh rows first (a degraded price, older than 48 h, is shown and flagged but never ranked first while a fresh one exists), fresh on-demand before community/marketplace rows, then compute subtotal ascending; the summary names any cheaper fresh rate ranked lower. If no fresh price exists, stale_fallback=true and the summary says the top row is not a current price. You supply the duration: job_hours, OR work_units plus throughput_units_per_gpu_hour from your own benchmark (job_hours = work_units / (throughput x gpus_per_job)). If you give workload + work_units but no throughput, a directly applicable published figure (closed-division MLPerf Inference, exact GPU variant, inference only) is used: throughput_basis='published_benchmark' with throughput_source {url, quote, date, benchmark, derived, caveats}; your throughput always wins. Otherwise not_found 'missing: throughput for {gpu}'. Whole-instance billing: instances = ceil(gpus_per_job / gpus_per_instance), all allocated GPUs billed (9 GPUs on an 8-GPU-only offer bill 16). Reserved and held-stale rates are never ranked; spot/interruptible only if allow_interruptible. Each result has …
| Parameter | Type | Description | |
|---|---|---|---|
gpu_family | string | optional | |
gpus_per_job | integer | optional | |
job_hours | number | optional | |
work_units | number | optional | |
throughput_units_per_gpu_hour | number | optional | |
workload | string | optional | |
allow_interruptible | boolean | optional | |
min_vram_gb | integer | optional | |
runs_per_month | number | optional | |
top_n | integer | optional | |
max_data_age_hours | number | optional | |
region | string | optional | |
gpu | string | optional | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "estimate_job_cost",
"arguments": {}
}
}get_dealread-onlyexternal accessGet one deal's public record by id, including redeem_url and agent_instructions.
Get one deal's public record by id, including redeem_url and agent_instructions.
| Parameter | Type | Description | |
|---|---|---|---|
id | string | required | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "get_deal",
"arguments": {
"id": "..."
}
}
}get_price_changesread-onlyexternal accessGPU rental price changes among tracked providers, each with old/new hourly price, change_pct, direction, the …
GPU rental price changes among tracked providers, each with old/new hourly price, change_pct, direction, the vendor source_url, a stable event_id (dedupe key) and rate_id. Marketplace median moves also have composition_flag (composition, price, mixed, unknown, or null) from successive median, minimum and listing count; composition, mixed and unknown are not price moves. since: 'YYYY-MM-DD' returns every change on or after that date, newest first (daily snapshots in /api/v1/gpu-prices/history.json); omitted returns the latest day's moves (/api/v1/gpu-price-changes.json). cursor: pass next_cursor from a previous call to get every change after it, oldest first, with has_more; loop until has_more is false, and store next_cursor only after processing (a watcher never misses or repeats events). error cursor_expired means re-sync with since and dedupe by event_id. kinds: optional subset of down, up, new, removed, same. gpu: optional model filter (substring, e.g. 'H100'). limit 1-200. Includes data_checked_at and next_refresh_utc (prices refresh daily at about 10:23 UTC).
| Parameter | Type | Description | |
|---|---|---|---|
since | string | optional | |
gpu | string | optional | |
limit | integer | optional | |
cursor | string | optional | |
kinds | string[] | optional | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "get_price_changes",
"arguments": {}
}
}list_watchesread-onlyexternal accessHosted only. Your watches and per-key alert counts.
Hosted only. Your watches and per-key alert counts.
| Parameter | Type | Description | |
|---|---|---|---|
api_key | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "list_watches",
"arguments": {}
}
}optimize_agent_stackread-onlyexternal accessEstimate a monthly stack cost from a short description of the agent being built. description: free text, e.g. …
Estimate a monthly stack cost from a short description of the agent being built. description: free text, e.g. 'scrapes 20k JS pages/month via 5 GB residential proxies, 3k search queries, summarizes with an LLM, budget $300/mo'. Parsed deterministically (keywords/regex, no LLM) into components: gpu, scraping, proxy, search, llm_api, email, hosting, database, monitoring. Optional fields override the parse: gpu_hours (GPU-hours/month), gpu_model ('H100', 'A100', 'RTX 4090'), model_size_b (billions of params; sets a minimum VRAM using FP16 x 1.2), scrape_pages, js_rendering, proxy_gb, search_queries, emails_sent, llm_api, monthly_budget_usd. Each component gets the cheapest option with a cited price among tracked providers, with calc, source_url, verified_at, referral_disclosure (set when the link is a referral link) and manual-signup notes. Components with no cited price say 'no sourced price tracked' (nothing is estimated). Totals show the math and say whether they are complete. Pulled listings are excluded. Every option has data_age_hours, degraded (> 48 h) and stale (> 7 days); GPU prices older than 168 h never rank by default (listed in the line's stale_prices); max_data_age_hours replaces that default and drops older options (excluded_too_old counts them). Same data as optimize_stack, compare_gpu_prices and the deal feed.
| Parameter | Type | Description | |
|---|---|---|---|
description | string | optional | |
workload | string | optional | |
gpu_hours | number | optional | |
gpu_model | string | optional | |
model_size_b | number | optional | |
scrape_pages | number | optional | |
js_rendering | boolean | optional | |
proxy_gb | number | optional | |
search_queries | number | optional | |
emails_sent | number | optional | |
llm_api | boolean | optional | |
monthly_budget_usd | number | optional | |
exclude_manual_signup | boolean | optional | |
max_data_age_hours | number | optional | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "optimize_agent_stack",
"arguments": {}
}
}optimize_stackread-onlyexternal accessCheapest suitable stack among tracked providers for monthly needs (set the quantities you need, rest 0). …
Cheapest suitable stack among tracked providers for monthly needs (set the quantities you need, rest 0). scrape_pages: pages/month without JS rendering. scrape_pages_js: pages/month needing JS rendering. proxy_gb: residential proxy GB/month. emails_sent: transactional emails/month. search_queries: search API queries/month. gpu_hours: GPU-hours/month (1 GPU), optionally gpu_model e.g. 'H100', 'A100', 'RTX 4090'. Each line has the chosen plan, monthly_cost_usd, calc, vendor quote, source_url, checked_at and assumptions, plus alternatives. A vendor price that isn't published is 'unknown' (never estimated). manual_signup_only=true means a human must sign up (never automate it); exclude_manual_signup=true drops those plans. Same data as https://thelanternworks.com/api/v1/stack-pricing.json and /tools/stack-optimizer/.
| Parameter | Type | Description | |
|---|---|---|---|
scrape_pages | integer | optional | |
scrape_pages_js | integer | optional | |
proxy_gb | number | optional | |
emails_sent | integer | optional | |
search_queries | integer | optional | |
gpu_hours | number | optional | |
gpu_model | string | optional | |
exclude_manual_signup | boolean | optional | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "optimize_stack",
"arguments": {}
}
}rebuild_quoteread-onlyexternal accessReopen a past quote. Pass the `receipt` object returned by optimize_stack, optimize_agent_stack or …
Reopen a past quote. Pass the `receipt` object returned by optimize_stack, optimize_agent_stack or estimate_job_cost. Returns receipt_intact (unkeyed sha256 matches), reproduced (true only if each line's subtotal recomputes AND its rate_id and price match the pinned feed revision or that date's published daily snapshot; "arithmetic_only" with provenance_verified=false when that can't be confirmed), per-line provenance, and revalidation against today's feed: per GPU line unchanged | price_changed | evidence_changed | held_today | not_in_feed_today (flat plans: not_revalidated), with revision 'same' or 'new'. Nothing is re-ranked.
| Parameter | Type | Description | |
|---|---|---|---|
receipt | string | required | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "rebuild_quote",
"arguments": {
"receipt": "..."
}
}
}search_dealsread-onlyexternal accessSearch agent work-stack deals (compute, search, email, unblock, hosting...). offer_type: verified_discount | …
Search agent work-stack deals (compute, search, email, unblock, hosting...). offer_type: verified_discount | listed_discount (not confirmed; verified_at null) | free_tier | trial | standard_referral | credit_only_referral. max_price excludes deals with unknown price. limit 1-25. Pass next_cursor back as cursor to page.
| Parameter | Type | Description | |
|---|---|---|---|
query | string | optional | |
category | string | optional | |
offer_type | string | optional | |
max_price | number | optional | |
tags | string[] | optional | |
limit | integer | optional | |
cursor | string | optional | |
mode | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "search_deals",
"arguments": {}
}
}watch_pricecan modifyexternal accessHosted only. Alert your webhook (signed POST) after the daily refresh when a tracked rate for `gpu` crosses …
Hosted only. Alert your webhook (signed POST) after the daily refresh when a tracked rate for `gpu` crosses at or below max_hourly_usd, or a change of a watched kind (down/up/new; default down) lands at or below it. Only ELIGIBLE rates match (same contract as compare/estimate): every word of gpu must match the model; 'A100 80GB' or vram_gb=80 never matches a 40GB row; tiers (on-demand, community, marketplace, interruptible/spot, dedicated-monthly, reserved; default all but reserved) never matches another class; a row confirmed sold out never matches (in_stock='require' also skips unknown stock). region: code from availability.available_regions; a rate with no unexpired region evidence stays unresolved and never matches. Each event is delivered at most once per watch (dedupe on event_id). Pass api_key or an Authorization: Bearer header.
| Parameter | Type | Description | |
|---|---|---|---|
gpu | string | required | |
max_hourly_usd | number | required | |
region | string | optional | |
vram_gb | integer | optional | |
tiers | string[] | optional | |
in_stock | string | optional | One of: exclude_sold_out, require |
kinds | string[] | optional | |
api_key | string | optional |
Example call (required arguments only, placeholder values):
{
"method": "tools/call",
"params": {
"name": "watch_price",
"arguments": {
"gpu": "...",
"max_hourly_usd": 0
}
}
}Badge
Maintain this server? Put its live status in your README:
[](https://mcp.aianytime.net/s/lantern-works/)Edit this entry · Report a problem
More in Cloud, ops & data
Checks record what a server answered, not whether it is safe. A listing is not a security audit; read a server's source and permissions before you connect it. Add a server · JSON · llms.txt · llms-full.txt
Browse: Agents, IDEs & developer tools · Search & discovery · Knowledge, memory & context · Media & generation · Security · Finance & markets · Marketing, content & social · Cloud, ops & data · Commerce, travel & logistics · Work & productivity · No account needed · Local servers