← All free tools · 🧩 Open Source
vLLM
✓ Verified against the official pageInference server with PagedAttention: maximum GPU throughput.
Free limit
Free limit
Open code: self-host free no expiry
Access type
Open source
Period
no expiry
Watermark
Not verified
Commercial use
Yes
Credit card
No
Sign-up
No
API
Yes: OpenAI-compatible API server, plus Anthropic Messages API and gRPC support; runs on your own hardware
Resolution and duration limits
—
Regional restrictions
Not specified
Checked
Nov 22, 2026
vLLM is an open-source LLM inference/serving library under Apache-2.0, free to install locally (uv/pip); the page lists no pricing tiers or usage quotas. Any limits come from your own hardware.
What the tool does
OpenAI-compatible server, batching, quantization, speculative decoding.
Official source
Apache-2.0 license
- github.com/vllm-project/vllm — Nov 22, 2026 (8149 симв.)
Free-tier change history
| Field | Before | After | Date |
|---|---|---|---|
| No changes recorded yet. | |||