Models
PEKPIK LLM exposes many chat, code, embeddings, image, video and audio models through an OpenAI-compatible surface. The live model catalog is available on the model section of the website.
List models via the API
Section titled “List models via the API”The standard OpenAI models endpoint returns everything your key can call:
curl https://aiapiv2.pekpik.com/v1/models \ -H "Authorization: Bearer $PEKPIK_KEY"Popular model IDs
Section titled “Popular model IDs”| Use case | Example model IDs |
|---|---|
| Flagship chat | claude-opus-4-7, gpt-5.4, gemini-2.5-pro |
| Fast general chat | deepseek-v3.2, glm-4.6, kimi-k2.5, gemini-2.5-flash |
| Reasoning | deepseek-r1-0528, kimi-k2-thinking |
| Code | codestral-latest |
| Embeddings | text-embedding-3-small, text-embedding-3-large |
| Images | gpt-image-2, gemini-2.5-flash image models |
| Audio | whisper-1 (transcription), TTS voices |
Genuine, full-fidelity models
Section titled “Genuine, full-fidelity models”The exact model and version you request is what runs. We never quantize, distill, or silently substitute a different model, and we continuously spot-check live outputs against each provider’s reference behaviour.
Usage and access
Section titled “Usage and access”Public keys are shared and temporary. Check the GitHub README for the latest published keys, use the playground to verify a key, and follow @getkeyway for fresh key drops and model alerts. See Usage & access.