424 models are now live on ApiLink
ApiLink's model catalog just went from a curated handful to 424 models across every major provider — synced automatically, kept up to date, and callable through the exact same OpenAI-compatible endpoint you're already using.
What changed
Previously, ApiLink sold a curated set of popular models. As of this release, we sync the full public model catalog directly from our upstream routing partner, which means:
- Every model from OpenAI, Anthropic, Google, Meta, Mistral, Qwen, DeepSeek, xAI, Cohere, and a couple dozen more providers is now available.
- New models show up automatically as providers release them — no waiting for us to manually add them.
- Pricing, context length, and modality metadata are pulled in with each model, not hand-entered.
| Before | Now | |
|---|---|---|
| Models available | A curated ~20-model shortlist | 424 |
| New model lag | Manual, days to weeks | Automatic, near-real-time |
| Coverage | Mainstream chat models only | Chat, vision, embeddings, and niche/regional models |
Using the new catalog
Nothing changes about how you call the API — same base URL, same API key, same OpenAI-compatible request format. Just pass a different model value. Browse the full, filterable list at apilink.io/models.
What's next
This is the foundation for a broader model-selection push — expect curated "best for X" collections (cheapest, fastest, best for agentic tool use) built on top of this catalog in a future update.