At a glance — Oct 2026
Updated 2026-10-27 · forward this page: /jtbds/cheap-gpu#briefScores as of 2026-09-14 (43 days old; Depot marks scores stale after 30 days). The pick is still Depot's current editorial pick; it has not been re-reviewed since. Prices below carry their own dates.
- Depot Pick
- Vast.ai · 7.20/10 for this job (curated estimate, medium confidence)
pricePerf primary; reliability heavily discounted for marketplaces.
- Runner-up
- RunPod · 8.34/10
Scores higher than the pick on: Capability fit, Latency / reliability, Developer & agent DX, Privacy / data retention, Maturity / track record.
- When not to use the pick
- Need SLA/security — RunPod Secure Cloud, Lambda, or CoreWeave. Never Vast for sensitive PII without review.
- Lock-in / migration
- The pick has lower stored lock-in risk than runner-up RunPod (2.5 vs 3.5).
- Sourced price (usd_per_gpu_hour)
Pick: no sourced price in this job's unit (no_sourced_price).
Cheapest matching SKU: SaladCloud RTX 3090 24GB (Lowest priority, batch) $0.09 — source, as of 2026-10-19. Not the pick; not quality-adjusted.
Price history: first seen 2026-10-19 · last changed no change recorded · history JSON
- Free option
- No free resource lists this job. A cheap paid SKU is not a free option.
- Failure modes to watch
- Partial tool success: Require a per-item status and a failure count. Success for the task requires the failed count to be zero or explicitly handed to a retry queue.
- Runaway spend: Set a budget and a tool-call cap per task. Alert when spend jumps. Stop the run at the cap instead of summarizing a partial trace as success.
Cite: Agent Provider Depot, Depot Pick for cheap-gpu: vast-ai, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=cheap-gpu
All sourced prices, cheapest first
Vast.ai
pricePerf primary; reliability heavily discounted for marketplaces.
Pricing (directional): As of 2026-09-14: P2P GPU marketplace; H100 commonly ~$1.5–2.5/hr class depending on host/interruptibility; prices fluctuate by supply. Interruptible instances cheapest; verify host reliability and security posture. Never default for regula…
Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.
Cheapest GPUs if your job can die and restart. Never put customer PII training here without eyes wide open.
Quote for Cheapest on-demand GPUs
Depot Pick stays Vast.ai. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=cheap-gpu
Pick is not a price winner. Cheapest matching USD per GPU-hour among candidates is salad at 0.09. The pick has no matching sourced price. Chat input $/1M was not borrowed.
Sourced price unavailable for the Depot Pick. No row in the pricing index.
9 of 14 candidates have no sourced price row. 9 are excluded from the USD per GPU-hour comparison (gpu_hour). min sourced GPU-hour SKU among candidates; not coerced to $/1M; per-second GPU was not included; not quality-adjusted
Sourced USD per GPU-hour for Cheapest on-demand GPUs
Candidates for this job that already have a matching sourced price (gpu_hour), cheapest first. As of 2026-10-27. Each figure is one SKU in that unit, not a blend of input and output, not a cross-modal rate, and not a Depot score. min sourced GPU-hour SKU among candidates; not coerced to $/1M; per-second GPU was not included; not quality-adjusted The Depot Pick is not a price winner.
| Provider | USD per GPU-hour | SKU | Source | As of |
|---|---|---|---|---|
| SaladCloud | $0.09usd_per_gpu_hour | RTX 3090 24GB (Lowest priority, batch) | pricing page | 2026-10-19 |
| RunPod | $0.34usd_per_gpu_hour | RTX 4090 Pod (Community Cloud) | pricing page | 2026-10-19 |
| Massed Compute | $0.97usd_per_gpu_hour | L40S (48GB) x1 | pricing page | 2026-10-19 |
| Verda | $1.845usd_per_gpu_hour | 1x L40S 48GB | pricing page | 2026-10-27 |
| Oracle Cloud GPU | $2usd_per_gpu_hour | VM.GPU.A10.1 | pricing page | 2026-10-19 |
Runners-up
RunPod
Depot default for self-serve GPUs. Secure Cloud for anything customer-facing; Community/spot for batch.
Oracle Cloud GPU
Secret weapon for cheap cloud GPUs if you tolerate OCI quirks. Not the agent-native DX winner.
FluidStack
Useful opportunistic H100 capacity if the live console quote clears ~$2.1–2.5/hr. Remain medium until a…
Thunder Compute
Indie-friendly A100/H100 when you want simpler than Vast marketplace. RunPod/Salad for interruptible fleets…
Verda
Top EU list-price neocloud peer to Nebius. Prefer Nebius Depot default for global neocloud; Verda when EU…
When NOT to use the top pick
Need SLA/security — RunPod Secure Cloud, Lambda, or CoreWeave. Never Vast for sensitive PII without review.
Full ranking (score table)
Ranked with this JTBD's weights. Scores are curated estimates unless marked otherwise. Agents: prefer /api/decide then this table for tie-breaks.
| # | Provider | Overall | Fit | $/perf | Latency | DX | As of |
|---|---|---|---|---|---|---|---|
| 1 | RunPod | 8.3 | 8.8 | 9.0 | 8.0 | 8.8 | 2026-10-19 |
| 2 | Verda | 8.2 | 8.6 | 8.2 | 8.4 | 8.4 | 2026-09-14 |
| 3 | Hyperstack | 8.0 | 8.2 | 8.4 | 8.0 | 8.0 | 2026-09-14 |
| 4 | Oracle Cloud GPU | 7.7 | 8.0 | 8.8 | 8.0 | 6.5 | 2026-09-14 |
| 5 | Thunder Compute | 7.7 | 7.2 | 8.8 | 7.4 | 8.2 | 2026-09-14 |
| 6 | Paperspace (DigitalOcean) | 7.6 | 7.2 | 7.8 | 7.8 | 8.2 | 2026-09-14 |
| 7 | FluidStack | 7.5 | 8.0 | 8.4 | 7.2 | 6.8 | 2026-09-14 |
| 8 | Massed Compute | 7.5 | 7.6 | 8.2 | 7.4 | 7.2 | 2026-09-14 |
| 9 | Oblivus | 7.3 | 7.4 | 8.4 | 7.0 | 7.2 | 2026-09-14 |
| 10 | TensorDock | 7.2 | 7.2 | 8.8 | 6.2 | 7.2 | 2026-09-14 |
| 11 | Vast.aiDepot Pick | 7.2 | 7.0 | 9.8 | 4.5 | 7.2 | 2026-09-14 |
| 12 | CloudFerro | 7.2 | 7.2 | 5.8 | 8.4 | 6.8 | 2026-09-14 |
| 13 | Shadeform | 7.2 | 7.0 | 8.4 | 6.8 | 8.0 | 2026-09-14 |
| 14 | SaladCloud | 7.0 | 6.5 | 9.5 | 5.0 | 7.0 | 2026-10-19 |
Score methodology for this job
pricePerf primary; reliability heavily discounted for marketplaces.
capabilityFit: 22%pricePerformance: 25%latencyReliability: 15%developerDx: 12%privacyPosture: 8%lockInRisk: 8%maturity: 10%