Find the right provider to self-host your LLM
Compare LLM hosting plans, open models and the experts who deploy them.
Latest LLM hosting plans
All plans →| Plan | LLM host | Type | Pricing range | Model families |
|---|---|---|---|---|
| Fireworks On-Demand Deployments | Fireworks AI | Dedicated endpoint | US$8 – US$20Per GPU-hour, Reserved or committed capacity | Llama, Mistral, DeepSeek, Qwen, Gemma, gpt-oss |
| Fireworks Serverless | Fireworks AI | Serverless API | US$0.05 – US$22.5Pay per token | Llama, Mistral, DeepSeek, Qwen, Gemma, gpt-oss |
| Cloudflare Workers AI | Cloudflare | Serverless API | From US$0.045Pay per token | Llama, Mistral, Qwen, gpt-oss |
Latest LLM models
All models →gpt-oss-120b
OpenAI's larger open-weight reasoning model; runs on a single 80 GB GPU.
- Parameters
- 117B total, 5.1B active (mixture of experts)
- Context
- 128K tokens
Qwen2.5 72B Instruct
The largest Qwen2.5 model, strong at general tasks and multilingual work.
- Parameters
- 72.7B
- Context
- 128K tokens
DeepSeek-V3
DeepSeek's large general-purpose mixture-of-experts model.
- Parameters
- 671B total, 37B active (mixture of experts)
- Context
- 128K tokens
Latest news
All news →LLM Hosting Roundup: Anthropic’s capacity crunch, Vera Rubin and DeepSeek’s $15B (23 Oct 2026)
Anthropic's hunt for cloud capacity, NVIDIA's DSX POD and new Vera Rubin inference benchmarks, Dell's first Groq 3 rack, DeepSeek's reported $15B raise and Reflection's open-weights Beam model.
Fireworks AI: serverless and on-demand hosting now in our directory
We've added Fireworks AI, the fast open-model inference platform founded by Meta's former PyTorch leads, with its serverless and on-demand hosting plans and 17 supported models.
US Homeland Security bans DeepSeek from federal civilian agencies
A new DHS order bars US civilian agencies from buying DeepSeek products and tells them to remove the Chinese lab's AI tools from their systems. It also covers third-party products built on DeepSeek.
