Qwen3 8B
qwen3:8b
General chat, coding and structured extraction
Free to use until Dec 1, 2026
Use modelModels
Four models. Four modalities.
Text, vision, image generation and speech. Each model has a native route — free and open for production until Dec 1, 2026.
Text
Qwen-VLM
qwen-vlm
Images, screenshots and document understanding
Coming soon
Not yet availableFlux Schnell
flux-schnell
Fast text-to-image generation
Coming soon
Not yet availableWhisper
whisper
Speech transcription and translation
Coming soon
Not yet availableNo account, no key, no cost — until Dec 1, 2026.
One API, native routes
No special cases. No new SDK.
Text and vision stream
Token-by-token output for chat and visual understanding.
Tools and JSON
Structured output and function calling without a custom SDK.
Purpose-built routes
Chat, image generation and transcription each get a native endpoint.