{"id":"online.pangle.compute/gpu-rental-market","slug":"online-pangle-compute-gpu-rental-market","title":"compute.pangle.online — GPU rental market","description":"Live GPU rental market: 2,500+ offers across a dozen provider feeds. History, watches, limit orders.","version":"1.2.0","publisher":{"kind":"domain","id":"compute.pangle.online","label":"compute.pangle.online","url":"https://compute.pangle.online"},"website":"https://compute.pangle.online/agents/","repository":"https://github.com/aitools420/compute-wick","remotes":[{"type":"streamable-http","url":"https://compute.pangle.online/mcp/"}],"packages":[],"published":"2026-08-28","updated":"2026-08-28","status":"online","auth":"","latency_ms":1731,"checked":"2026-11-19","uptime":100,"checks":3,"tools":[{"name":"search_offers","title":"","description":"Search live GPU offers. gpu matches the model name (e.g. '4090', 'H100').\n    offer_class: on_demand | interruptible | reserved. country: ISO-2 (e.g. US),\n    a comma list ('DE,NL'), or 'EU' for the 27 member states. min_gpu_count\n    filters to multi-GPU configs (8 = full 8x nodes). region substring-matches\n    the provider's region/datacenter label. Returns offers ranked\n    cheapest-first by pr","read_only":false},{"name":"get_offer","title":"","description":"Fetch one offer by its id (from search_offers).","read_only":false},{"name":"market_stats","title":"","description":"Live market snapshot: total offers, GPU models, idle-capacity share,\n    per-provider freshness/health, cheapest price per model, current fee_bps.","read_only":false},{"name":"price_history","title":"","description":"Price history for one GPU model (exact name from search_offers/market_stats,\n    e.g. 'RTX 4090'). offer_class: on_demand | interruptible. Returns per-provider\n    series of [ts, min_price_per_gpu_hr, median_price_per_gpu_hr, offer_count];\n    ranges past 3 days are bucketed (hourly, then 6-hourly past a week).","read_only":false},{"name":"idle_history","title":"","description":"The idle-capacity index over time: share of listed GPU capacity that is\n    interruptible (spot/bid) — idle hardware looking for work. Points are\n    [ts, idle_share, idle_offers, total_offers].","read_only":false},{"name":"best_value","title":"","description":"Perf-per-dollar board: for each GPU model, best live price vs its FP16\n    tensor throughput (dense, vendor spec sheets) -> TFLOPS per dollar-hour.\n    The answer to 'most compute for my budget' rather than 'cheapest card'.\n    Models without a defensible public spec figure are listed as unrated.","read_only":false},{"name":"provider_reliability","title":"","description":"How dependable each provider's data feed has been for this station\n    (poll success over the trailing window), plus the provider's own average\n    machine reliability where reported. feed_score is about the DATA, not\n    their hardware.","read_only":false},{"name":"price_position","title":"","description":"Rent-now-or-wait context: where the current best price for one GPU\n    model sits inside its own trailing 7/30-day range (percentile, 24h trend,\n    verdict). Descriptive, not a forecast.","read_only":false},{"name":"spot_spread","title":"","description":"On-demand vs interruptible spread: the live discount for idle capacity\n    per model; pass gpu_model for its history too.","read_only":false},{"name":"will_it_fit","title":"","description":"Will this model fit, and what does a fitting machine cost right now?\n    Deterministic VRAM arithmetic — weights (params x bytes/precision) + KV\n    cache (2 x layers x kv_heads x head_dim x context x batch) + stated\n    overhead — checked against the live book, cheapest fitting offers first.\n    Pass a known model slug (llama-3.1-8b/70b/405b, mistral-7b, mixtral-8x7b,\n    qwen2.5-7b/32b/72b, gemm","read_only":false},{"name":"true_cost_notes","title":"","description":"The billing fine print behind the $/hr, verified per provider from\n    their own published pages (source URL + date on every claim): billing\n    granularity (per-second vs per-minute), whether storage bills while\n    stopped, bandwidth/egress fees, and the commonly-missed gotchas (e.g.\n    Vast bandwidth never appears in $/hr; RunPod stopped volume disk costs\n    double). Providers we have not ver","read_only":false},{"name":"register_account","title":"","description":"Create a metering account: returns a bearer token (shown once, we keep\n    only a hash). Pass it as account_token on rent calls to build a usage\n    ledger you can read back with account_usage. Optional — renting works\n    without one.","read_only":false},{"name":"account_usage","title":"","description":"Your metered usage ledger: placements, destroys, estimated hours and USD,\n    and the platform fee (2.5%, $0.01 min, $5.00 cap per lease per UTC day) priced\n    at the rate live at event time. Distinct from fee_bps, the display markup.","read_only":false},{"name":"create_watch","title":"","description":"Create a price watch (tripwire): fires when the best live fee-adjusted\n    price per GPU-hour matching gpu (and optional offer_class) drops to or\n    under max_price_per_gpu_hr. Checked every poll (~5 min). Returns the watch\n    with its id — the id is the only key; poll it with watch_status, or give a\n    public webhook_url to be POSTed on each trip. Watches re-arm when the\n    price climbs 2% ba","read_only":false},{"name":"watch_status","title":"","description":"Current state of a watch: armed/tripped, last price seen, recent\n    events, and the best matching offer right now.","read_only":false},{"name":"delete_watch","title":"","description":"Delete a watch by id.","read_only":false},{"name":"create_limit_order","title":"","description":"Place a standing LIMIT ORDER: when the best live fee-adjusted price for\n    gpu (and optional offer_class) trades at or under max_price_per_gpu_hr,\n    the station cuts a signed FILL TICKET naming the exact offer. The station\n    never holds your provider key — the ticket is executed by whoever does:\n    the open-source keyholder sidecar (/agents/#sidecar), your own agent\n    long-polling POST /ap","read_only":false},{"name":"limit_order_status","title":"","description":"State of a limit order (armed/ticketed/filled/cancelled/expired), last\n    price seen, recent events. Pass order_secret to also read the live fill\n    ticket when one is cut.","read_only":false},{"name":"cancel_limit_order","title":"","description":"Cancel a limit order. Requires the order_secret from create_limit_order.","read_only":false},{"name":"rent_x402","title":"","description":"Rent a prepaid block of GPU time with a wallet and NO account, NO API key\n    and no human, over x402 (HTTP 402 micropayments, USDC on Base).\n\n    Call it with no arguments and you get the payment challenge: an x402\n    PaymentRequirementsResponse whose accepts[0].extra.x402_fee states the block\n    price, the platform fee IN DOLLARS, that there are no refunds, and that you\n    are charged only af","read_only":false},{"name":"rent_offer","title":"","description":"Place a rental on YOUR provider account (BYO key — it is used for this\n    one call and never stored). Default is a DRY RUN returning exactly what\n    would be executed; a live placement needs dry_run=false AND confirm=true.\n    The offer is re-quoted live first and refused if its price exceeds\n    max_price_per_gpu_hr. Executes on vast and runpod offers. Pass an\n    idempotency_key so retries ret","read_only":false},{"name":"rent_best","title":"","description":"One-shot rent: search the live book with these filters and place on the\n    best (cheapest) matching offer — 'cheapest H100 under $2/hr, go'. Same\n    guarantees as rent_offer (dry-run default, live re-quote, your\n    max_price_per_gpu_hr is absolute). If the best offer is already gone it\n    walks to the next, at most 3, and reports what it skipped.","read_only":false},{"name":"rental_status","title":"","description":"Live state of an instance on your account: running/starting/stopped/\n    gone, uptime, price, estimated spend so far, and the budget guard's state\n    if one is armed. Read-only; your key passes through and is never stored.","read_only":false},{"name":"destroy_rental","title":"","description":"Destroy an instance previously placed on your account (vast or runpod).\n    Your key passes through and is never stored.","read_only":false},{"name":"fill_limit_order","title":"","description":"Execute a limit order's live fill ticket on YOUR provider key (used for\n    this one call, never stored). Only works while the order is ticketed; the\n    offer is re-quoted live and refused above the order's line. Idempotent per\n    order — a retry returns the first receipt rather than renting a second\n    machine. The order's auto_destroy_budget_usd (if set) arms the budget\n    guard exactly as r","read_only":false}],"server_info":{"name":"compute.pangle.online","title":"compute.pangle.online — GPU rental market","version":"1.2.0","websiteUrl":"https://compute.pangle.online/agents/"},"category":"real-estate","tags":["ecommerce","crypto","productivity"],"kind":"remote","source":"registry","submitted":{},"first_party":true,"domain_rank":null,"score":100,"indexable":true,"first_seen":"2026-08-28","stars":0,"badges":[],"sponsor":null,"page":"https://agentsindex.org/agents/online-pangle-compute-gpu-rental-market/"}