NVIDIA B300 price per GPU-hour in November 2026: OVHcloud vs AWS vs Scaleway, with commitments and egress
Blog post from Qovery
A comparison of provider price data dated November 15, 2026, finds Scaleway’s eight-GPU NVIDIA B300 node in Paris is the cheapest published option at €60 per hour, or €7.50 per GPU-hour, though its availability was marked “shortage.” AWS charges $142.42 per hour in us-east-1 and $217.54 in London on demand, while offering substantial commitment discounts: a three-year, all-upfront EC2 Instance Savings Plan lowers the us-east-1 rate to about $53.55 per hour. OVHcloud had announced B300 availability for November but had not published a price or region; an inactive, unlisted catalog entry showed €85.60 per hour, so the figure is provisional. The systems share an eight-GPU configuration but differ in host specifications and networking, with AWS offering a high-bandwidth fabric suited to multi-node workloads, while Scaleway and the listed OVHcloud configuration emphasize single-node use. Commitment plans at Scaleway and OVHcloud exclude GPU instances, and internet egress can affect costs: AWS charges for outbound data beyond a free allowance, whereas OVHcloud includes it in most regions and Scaleway lists no per-GB charge for instance traffic. The comparison therefore depends not only on hourly rates but also on capacity, workload scale, regional access, data transfer, and negotiated terms; it suggests Scaleway for available single-node European workloads, AWS for multi-node training or discounted long-term use, and waiting for OVHcloud’s confirmed pricing and regional details when its platform or included egress is important.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| Kubernetes | 11 | 3,388 | 294 | 94 | -34% |
| MCP | 10 | 7,404 | 711 | 183 | -38% |
| AI Agents | 1 | 4,689 | 923 | 218 | -33% |
| Platform Engineering | 1 | 1,080 | 220 | 61 | -46% |
| TPUs | 1 | 125 | 19 | 7 | -59% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.