1× RTX 3070 TI rentals from Vast.ai

1 provider with 4 offers. Sorted by most reliable. Updated 2026-11-13 03:16 UTC.

1 provider · 4 offers

By providerAll offers
Sort: Most reliable
Most reliableLowest priceMost GPUs freeBest overallClosest to my GPU countMost GPU memoryMost GPUs per offer

JSONMarkdown

Vast.ai

marketplace4 offers

★★★★★★★★★★ 5.0

From host reliability5.0

From$0.097per GPU-hour · median $0.163

4 offers in stock · 5 GPUs free99.6% host reliabilityIn stock 100.0% of 30 days
  • 1× GPU · 8 GBQuebec, CA
    1 GPU free now
    $0.097per GPU-hour · about $71/mo
    View deal
  • 2× GPU · 8 GBNorway, NO
    2 GPUs free now
    $0.162per GPU-hour · about $118/mo
    View deal
  • 1× GPU · 8 GBNorway, NO
    1 GPU free now
    $0.164per GPU-hour · about $120/mo
    View deal
  • 1× GPU · 8 GBMassachusetts, US
    1 GPU free now
    $0.189per GPU-hour · about $138/mo
    View deal

Where to find RTX-3070-TI

Every provider we track, by region. 2 of 36 providers have it in stock now.

ProviderAmericas2 in stock · from $0.097Europe2 in stock · from $0.058Asia-Pacific1 in stock · from $0.067Elsewhere1 in stock · from $0.067
Clore.ai★★★★★★★★★★ 4.4 (956)From $0.1034 in stockFrom $0.05820 in stockFrom $0.0671 in stockFrom $0.06714 in stock
Vast.ai★★★★★★★★★★ 5.0From $0.0972 in stockFrom $0.1622 in stockNot offeredNot offered

Not offered by: RunPod, Akash, Cudo Compute, io.net, Koyeb, Thunder Compute, Akamai Linode, Microsoft Azure, Oracle Cloud, AWS, OVHcloud, Lium, Exoscale, Nosana, GPU.ai, SimplePod, IMWT (via Shadeform), Massed Compute (via Shadeform), Lambda (via Shadeform), Crusoe (via Shadeform), Paperspace (via Shadeform), Hyperstack (via Shadeform), Boost Run (via Shadeform), Latitude.sh (via Shadeform), Scaleway (via Shadeform), Voltage Park (via Shadeform), Denvr (via Shadeform), Excess Supply (via Shadeform), Horizon (via Shadeform), Verda (via Shadeform), DigitalOcean (via Shadeform), Nebius (via Shadeform), Vultr, Phyntec (via Shadeform).

For agents

Free to use from code: a JSON API, an MCP server, and every page as Markdown. No API key.

curl "https://gpuscout.cc/api/v1/providers/ranking?gpu=h100&count=8"

llms.txt · OpenAPI · MCP endpoint: https://gpuscout.cc/mcp