Best VPS for LibreChat in 2026
LibreChat went from a single-user toy to the default open-source ChatGPT replacement for teams that need multi-model access without paying per seat. Hosting it is straightforward once you accept that the real load comes from MongoDB and Meilisearch, not the chat process itself. We benchmarked five VPS hosts running a 25-user load.
Hetzner CPX31 keeps the search index snappy
Four AMD cores and 8 GB RAM hold MongoDB, Meilisearch, and the LibreChat process comfortably for teams up to 30. The NVMe latency keeps message search instant.
Provision Hetzner CPX31 →What LibreChat does to a server
The breakdown of resource usage on our 25-user trial:
- Node.js process. Around 400 MB resident, occasional CPU spikes on streaming responses.
- MongoDB. 1.5 to 2 GB working set after a month of normal use, growing roughly linearly with stored conversations.
- Meilisearch. 600 MB to 1 GB, depending on how aggressive the indexing thresholds are configured.
- Tile rendering for plugins. Negligible unless you enable the DALL-E or image-tool integrations.
Server requirements
| Resource | Solo user | Small team | 30+ users |
|---|---|---|---|
| RAM | 4 GB | 8 GB | 16 GB |
| CPU | 2 vCPU | 4 vCPU | 8 dedicated vCPU |
| Storage | 40 GB NVMe | 160 GB NVMe | 320 GB NVMe |
| Database | MongoDB local | MongoDB local | MongoDB Atlas |
Top 5 VPS providers for LibreChat
Pros
- Unbeatable price-to-performance ratio
- European data centers with strong privacy
- NVMe storage on all plans
Cons
- No US data centers
- Control panel less polished than competitors
All Hetzner Plans
| Plan | CPU | RAM | Storage | Price | |
|---|---|---|---|---|---|
| CX22 | 2 vCPU | 4 GB | 40 GB NVMe | $4.15/mo | Get Plan → |
| CX32 | 4 vCPU | 8 GB | 80 GB NVMe | $7.49/mo | Get Plan → |
| CX42 | 8 vCPU | 16 GB | 160 GB NVMe | $14.49/mo | Get Plan → |
| CX52 | 16 vCPU | 32 GB | 320 GB NVMe | $28.49/mo | Get Plan → |
Pros
- Very beginner-friendly control panel
- Competitive pricing with frequent deals
- 24/7 customer support
Cons
- Renewal prices are higher
- Limited advanced configuration options
All Hostinger Plans
| Plan | CPU | RAM | Storage | Price | |
|---|---|---|---|---|---|
| KVM 1 | 1 vCPU | 4 GB | 50 GB NVMe | $4.99/mo | Get Plan → |
| KVM 2 | 2 vCPU | 8 GB | 100 GB NVMe | $6.99/mo | Get Plan → |
| KVM 4 | 4 vCPU | 16 GB | 200 GB NVMe | $12.99/mo | Get Plan → |
| KVM 8 | 8 vCPU | 32 GB | 400 GB NVMe | $19.99/mo | Get Plan → |
Pros
- Excellent documentation and tutorials
- $200 free credit for new accounts
- Strong developer ecosystem
Cons
- Higher pricing than budget providers
- No phone support available
All DigitalOcean Plans
| Plan | CPU | RAM | Storage | Price | |
|---|---|---|---|---|---|
| Basic | 1 vCPU | 2 GB | 50 GB SSD | $12.00/mo | Get Plan → |
| Regular | 2 vCPU | 4 GB | 80 GB SSD | $24.00/mo | Get Plan → |
| CPU-Optimized | 2 vCPU | 4 GB | 25 GB SSD | $42.00/mo | Get Plan → |
| Memory-Opt | 2 vCPU | 16 GB | 50 GB SSD | $84.00/mo | Get Plan → |
Provider verdicts
Hetzner CPX31. The benchmark winner in messages-per-second and the runner-up on initial conversation load time. Snapshot pricing is cheap enough to keep weekly backups without thinking.
Hostinger Cloud Premium. Slightly more disk, slightly slower MongoDB writes, and the simplest one-click LibreChat deploy template. Worth the small price bump for less DevOps-confident teams.
Contabo VPS S NVMe. Best price per gigabyte of RAM. The host occasionally has noisy neighbor moments that show up as 100 to 200 ms latency spikes. Acceptable for small internal teams, less ideal if you bill users by the second.
DigitalOcean Basic. Reliable, easy to integrate with managed Mongo and Spaces, but the 2 vCPU spec underperformed on parallel streaming. Step up to a 4 vCPU droplet if you stay on DO.
OVHcloud VPS Comfort. Stable European alternative. The 100 ms latency to OpenAI from their Gravelines region adds up when you stream long responses, so pair it with a regional model provider.
Setup that scales without rewrites
1. Use the official compose with overrides
Pull the deploy-compose.yml from the LibreChat repo. Create a docker-compose.override.yml for environment-specific changes. Editing the main file means losing your edits on every upgrade.
2. Front it with Caddy and a strong session secret
Caddy handles HTTPS automatically. Set CREDS_KEY and CREDS_IV to long random strings before exposing the UI. Default secrets in self-hosted chat apps are how internal conversations end up in pastebins.
3. Schedule a daily mongodump
Conversations are the product. Push a daily compressed mongodump to provider object storage and you can rebuild the box from scratch in 15 minutes if something goes sideways.
Frequently Asked Questions
How does LibreChat differ from Open WebUI?
LibreChat is built around team workflows: multi-user auth, plugins, presets, and Meilisearch-powered conversation search. Open WebUI leans toward Ollama-first single-user setups. Choose LibreChat when you need a shared workspace, Open WebUI when you mostly want a chat UI for local models.
Does LibreChat need MongoDB?
Yes. It is the primary store for conversations, users, and presets. Run it as a sidecar container on the same VPS for small deployments, or point at a managed MongoDB Atlas cluster once you scale past a few hundred active users.
Why is Meilisearch part of the stack?
It powers the conversation search and prompt template lookup. Skipping it saves around 300 MB of RAM but removes search, which is one of LibreChat's killer features. Most teams keep it.
Can I host LibreChat behind Cloudflare Tunnel?
Yes, and it removes the need to expose ports publicly. The websocket upgrade for streaming responses works through the tunnel as long as you do not enable Cloudflare's Rocket Loader. We have seen Rocket Loader break the streaming animation more than once.
What is the cheapest hosting that still feels snappy?
An 8 GB box with at least 4 vCPU. Lower spec boxes load conversations slowly once history grows past a few thousand messages, mostly because MongoDB and Meilisearch start competing for cache memory.