Tier lists / LLMs & model APIs
🧠 Tier list · 13 tools
LLMs & model APIs
Frontier and open models, routers and inference providers. Ranked with our 100-point rubric. Scores are initial editorial estimates.
Featured (paid) Featured spot available. It's labeled and never affects the tier list. Get featured →
S85–100
A70–84
B55–69
No tools in this tier yet.
C0–54
No tools in this tier yet.
Pick this if…
| Tool | Tier | Use it when… |
|---|---|---|
| OpenAI API 88/100 | S | you want the broadest model lineup and ecosystem. |
| Anthropic Claude API 87/100 | S | you build coding agents or need top reasoning and tool use. |
| Google AI Studio / Gemini API 87/100 | S | you want long context, multimodal input and a strong free tier. |
| Claude Haiku 5.5 86/100 · provisional | S | you run many agent sub-tasks and the token bill hurts. |
| Ollama 84/100 | A | you want private local models with one command. |
| OpenRouter 83/100 | A | you want to switch models without rewriting code. |
| Qwen 79/100 | A | you want strong open weights in many sizes. |
| DeepSeek 76/100 | A | you want strong reasoning at very low cost or to self-host. |
| Grok API (SpaceXSI / x.ai) 75/100 | A | you want Grok models in your app. |
| LM Studio 75/100 | A | you want local models with a friendly GUI. |
| Mistral AI 73/100 | A | you want strong open-weight options or EU hosting. |
| Groq 72/100 | A | latency matters, e.g. voice agents. |
| Together AI 71/100 | A | you want hosted open models and fine-tuning. |