Baseten
Free tierThe fastest AI model inference platform for production-grade deployments
Key strengths
Facts verified against baseten.co on
Recent changes
- Free tier: now available
- Plans updated: Basic plan is $0/month pay-as-you-go; Pro plan requires a quote with volume discounts; Enterprise plan requires a quote. Compute billed per minute (e.g. H100 80 GiB at $0.10833/min). Model APIs priced per 1M tokens.
Get alerts when Baseten changes
Free. We'll email you when pricing, the free tier, API access or open-source status changes.
Baseten is a high-performance model inference platform built for AI/ML engineers deploying custom, fine-tuned, or open-source models in production at scale. Founded in 2022, it offers a usage-based pricing model starting at $0/month (pay-as-you-go), with Pro and Enterprise tiers available via quote. Compute is billed per minute (e.g., H100 80 GiB at $0.10833/min), and model APIs are priced per 1M tokens. Key technical capabilities include low-latency, high-throughput inference; fast cold starts; 99.99% uptime SLA; cross-cloud and self-hosted deployment options; and specialized optimizations for LLMs, image generation, transcription, TTS, and embeddings. Baseten exposes a REST API and integrates with tools including Datadog, Vercel, Canva, Cursor, Notion, and ComfyUI. The platform is open-source with 1,220 GitHub stars and supports self-hosting.