We have just added 50 media models to YepAPI. Each one ran a real job through the API before going live. They use the same `POST /v1/media/queue` endpoint, the same job polling and the same API key as every other media model, so switching is a one-line change to `model`.
What's new
- ElevenLabs on YepAPI: Eleven v4, v4 Turbo, v3, v3 Conversational, Multilingual v2, Flash v2.5 / v2 and Turbo v2.5 / v2, from $0.0844 per 1,000 characters
- GPT Image 2.5 Sunburst and Flare from $0.02/image, FLUX 3 Image from $0.11, FLUX 3 Image Klein from $0.02
- Seedream 5.0 Pro, Lite and Flash; Qwen Image 3, 3 Pro, 2.1 Pro and 2.1 Turbo; MAI-Image-2.6 and 2.6 Flash; Grok Imagine Image 2.0; Meta Muse Image; Tencent Hy Image 3.5
- Image editing models Krea K3 Edit and Recraft V4 Styles (four variants) take your image in `imageData`
- Video: Seedance 2.5, Seedance 2.0 Mini, Wan 3.0 and Wan 3.0 Prime, Hailuo 3 Max, FLUX 3 Video (up to 4K), HeyGen Video and Grok Imagine Video 1.5 Lite from $0.04/sec
- Voices: Gemini 3.8 Flash TTS and Flash Lite TTS, MAI-Voice-2.1 and 2.1 Flash, Deepgram Flux TTS, Voxtral TTS 3, HeyGen Voice and Voice Turbo, Fish Audio Drama 3 and Seed Audio 1.0
- Lyria 3.5, Google DeepMind's newest music model, at $0.16 per track
- Video `options.duration` now accepts up to 30 seconds for models that support it (Wan 3.0, Seedance 2.5)
Try one
Every new model works with the existing media queue. Submit a job, then poll `GET /v1/media/status/{jobId}`.
curl -X POST https://api.yepapi.com/v1/media/queue \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "elevenlabs/eleven-v4", "prompt": "Welcome back! Here is your daily briefing.", "options": {"voice": "sarah"}}'PCM voices
HeyGen Voice and the Gemini 3.8 TTS models return raw PCM only. The result's `mimeType` now includes the sample rate, for example `audio/pcm;rate=24000;channels=1`, so you can wrap it in a WAV header without guessing.
See every media model
Image, video, music, speech and analysis models with their current prices.
Browse all AI models