Text generation

ModelMakerStatusSource
gemma-4-31B-itGoogleActivegoogle/gemma-4-31B-it
GLM-5.3-FlashZ.aiActivezai-org/GLM-5.3-Flash
Mistral-Small-3.2-24B-Instruct-2506Mistral AIActivemistralai/Mistral-Small-3.2-24B-Instruct-2506
Qwen3.8-27B-FP8QwenPreviewQwen/Qwen3.8-27B-FP8
Kimi-K3Moonshot AIActivemoonshotai/Kimi-K3

Embeddings

ModelMakerStatusSource
multilingual-e5-large-instructintfloatActiveintfloat/multilingual-e5-large-instruct
multilingual-e5-largeintfloatActiveintfloat/multilingual-e5-large

Reranking

ModelMakerStatusSource
bge-reranker-v2-m3BAAIActiveBAAI/bge-reranker-v2-m3

Speech to text

ModelMakerStatusSource
kb-whisper-largeKBLabActiveKBLab/kb-whisper-large
nb-whisper-largeNbAiLabActiveNbAiLab/nb-whisper-large
faster-whisper-large-v3SystranActiveSystran/faster-whisper-large-v3
pianissimoKlangActiveKlangAI/pianissimo-sv

Every model has a list price per million tokens, or per 1 000 minutes of audio, billed per second. Preview models may change before they become active.