Diffusion-based LLMs (Mercury) that generate tokens in parallel for faster, cheaper inference than autoregressive models.
Best AI LLMs & Foundation Models of 2026
All 90 LLMs & Foundation Models tools that are actually alive — cross-checked against multiple directories, liveness-verified, and ranked by real traffic. Dead links and clones removed. Verified means we checked — not that someone paid.
See the ranked list ↓Narrow it down
2 subcategoriesLLMs & Foundation Models splits into more focused areas — jump straight to the one you need.
The ranked list
Showing 18 of 90 live tools · ranked by real traffic
LLM Labs helps developers compare language models side-by-side for informed decisions.
Popular tool for downloading and running open-weight LLMs locally, with an optional paid cloud for larger models.
Earth Species Project is a nonprofit building AI models to decode animal communication across species, not a commercial product.
One API and SDK to reach 400+ AI models across providers, with routing, RAG and agents and no vendor lock-in.
Free, open-source desktop GUI for running local language models on your own computer via llama.cpp.
On-device AI chat app for Apple devices that runs open-source LLMs fully offline with no subscription, cloud calls, or account required.
Single frontier reasoning model behind one OpenAI-compatible endpoint, billed pay-per-token.
Builds physics-based foundation models and an agent for real-world prediction, used by energy traders for atmospheric forecasting.
A multimodal language model that outperforms GPT-4 and DeepSeek on major benchmarks. Benefit from an optimized MoE architecture that reduces computing costs by 30% while maintaining outstanding performance
Meta AI research model that binds six data modalities into one shared embedding space for cross-modal tasks.
Side-by-side comparison of ChatGPT, Claude, Gemini, and 40+ AI models on price, speed, and output to pick the best fit.
Arcee's open-weight (Apache 2.0) reasoning LLM tuned for long-horizon agents and tool use; ranks #2 on PinchBench and is cheap via API.
Platform for connecting clients to LLMs with zero backend and enhanced security.
Creative workspace plus API to generate images, video, music and text from many top models, billed via pay-as-you-go credits.
Alibaba's Qwen3.6-35B MoE model uses ~3B active params, tuned for agentic coding, vision, and long context.
French AI lab offering open and commercial LLMs plus a platform to build, customize and run AI agents and apps.
A web app for running and comparing side-by-side evaluations of multimodal AI models across reasoning and vision tasks.
LLMs & Foundation Models tools compared
| Tool | Best for | Free tier | Price | Monthly visits |
|---|---|---|---|---|
| Mercury by Inception Labs | Diffusion-based LLMs (Mercury) that generate tokens in parallel for faster, cheaper inference than autoregressive models. | — | Paid | — |
| LLM Labs | LLM Labs helps developers compare language models side-by-side for informed decisions. | — | — | — |
| Ollama | Popular tool for downloading and running open-weight LLMs locally, with an optional paid cloud for larger models. | Yes | $20/mo | — |
| Earth Species Project | Earth Species Project is a nonprofit building AI models to decode animal communication across species, not a commercial product. | Yes | Free | — |
| AnyAPI | One API and SDK to reach 400+ AI models across providers, with routing, RAG and agents and no vendor lock-in. | Yes | $19.9/mo | — |