Koyeb Serverless Platform
Serverless GPU / functions
Provided by Koyeb SAS
Offering overview
Push a container or Git repo and let the platform handle autoscaling, HTTP/2, WebSockets and gRPC routing: that workflow now extends to GPU-backed inference and fine-tuning jobs. Services scale down to zero, deploy with zero downtime, and can attach NVMe volumes and a serverless Postgres with pgvector for retrieval use cases. Accelerator choices include RTX 4000 SFF Ada, L4, RTX A6000, L40S, A100, RTX PRO 6000, H100, H200 and B200 in one- to eight-GPU shapes, plus Tenstorrent N300s on request. One-click deploys cover popular AI models and stacks, and Koyeb Sandboxes give agents isolated execution environments. The company is based in Boulogne-Billancourt near Paris and has announced it is joining Mistral AI. Usage-based pricing starts at $0.50 per hour for an RTX 4000 SFF Ada and $2.50 per hour for a single H100; CPU instances begin at $0.0022 per hour. Pro and Scale plans unlock seven regions across the US, EU and Asia, while enterprise customers get 50 on-demand regions, AWS regions and BYOC. Serverless Postgres is hosted in Washington D.C., Frankfurt and Singapore.
- Management
- Not stated
- Infrastructure
- Not stated
- Pricing model
- Not stated
- Support
- Not stated
- SLA
- Not stated
- Best for
- AI Applications, Model Inference, AI Agents, Fine-Tuning, Batch Processing, Vector Search
Technical specifications
- Containers
- Unknown
- Kubernetes
- Unknown
- Root access
- Unknown
- API access
- Unknown
- Autoscaling
- Unknown
- Persistent processes
- Unknown
AI capabilities
- Model serving
- Not stated
- AI frameworks
- Not stated
- Inference endpoints
- Unknown
- Fine-tuning
- Unknown
- Vector database
- Unknown
- Model catalog
- Unknown
- Monitoring
- Not stated
- Evaluation
- Not stated
Security & compliance
No controls are stated.
Geography
Provider headquarters are separate from offering infrastructure and availability.
Infrastructure regions
Germany, Europe · Singapore, Asia · United States, North America
Data-residency jurisdictions
Not stated
Service availability
Not stated
Support coverage
Not stated
Available plans
Pricing can change. Promotional prices are shown separately from standard prices.
| Plan | Price | Compute | Storage | GPU | Price checked | Actions |
|---|---|---|---|---|---|---|
| Koyeb Serverless PlatformNot stated | $0.50Usage Only | Unknown — RAM | Unknown | No | 7 October 2026 |
Questions & answers
- How are applications deployed on Koyeb?
- Users push a container or Git repository, and Koyeb handles autoscaling and HTTP/2, WebSockets and gRPC routing. Services support scale-to-zero and zero-downtime deployment.
- What AI workloads does Koyeb support?
- Koyeb supports GPU-backed inference, fine-tuning jobs and one-click deployment of AI models and stacks. Koyeb Sandboxes provide isolated execution environments for agents.
- What storage options does Koyeb provide?
- Services can attach NVMe volumes and serverless Postgres with pgvector for retrieval use cases.
- Which accelerators does Koyeb support?
- Documented NVIDIA options include RTX 4000 SFF Ada, L4, RTX A6000, L40S, A100, RTX PRO 6000, H100, H200 and B200 in one- to eight-GPU configurations. Tenstorrent N300s are offered on request.
Data sources & provenance
Where this information came from, how it was checked and when it was last looked at. “Last updated” records a change to the record — it does not mean the record was verified.
- Data quality
- Data quality: Unverified
- Primary source
- koyeb.com
- Sources cited
- 2
- Last reviewed
- 7 October 2026
- Last updated
- 18 October 2026