Added Qwen3.8 Max (0902), Qwen3.8 2.4T A95B, Qwen3.8 Flash and Qwen3.8 Omni Flash.
Added GLM 5.3 and GLM 5.3 Flash.
Added MiniMax H3 and MiniMax H3 Max for video generation.
Added Muse Spark 1.3.
Removed DALL-E and routed its requests to GPT Image 2.5 Flare.
Removed the Qwen 3.7 models.
Removed Stability AI services and routed requests to GPT Image 2.5 Flare.
Removed Llama Guard and updated moderation to use Mistral Moderation 2.
Updated Gemini 3.1 Flash TTS and Chirp 3 requests to use the new TTS models.
Updated old model IDs to use the closest current model and show the served model in the response.
Changed the name of Gemini 3.1 Flash Image to Nano Banana 2.
Changed the default chat model from Kimi K3 to Gemini 3.8 Flash with medium thinking.
Pricing
Optimized pricing across all models.
Lowered prices for many models.
Privacy
Added privacy labels to every model: End-to-end ZDR and SilentAPI protected.
Tools & Services
Added image compression and image conversion.
Added metadata removal for images, PDF files and Office files.
Added OCR and Mistral OCR 4.1.
Added Google Cloud Vision, including Document Text Detection for OCR.
Added Mistral Moderation 2.
Added option handling for media and reasoning requests. When a model does not support an option, SilentAPI converts the output or uses the closest supported value.
Improved IP lookup with SilentAPI's own data. IP data now updates daily.
Improved email lookup with real top-level domain and mail server checks. Email lists now update daily.
Platform
Moved hosting to new infrastructure for lower latency.
Significant backend optimizations to lower latency and improve stability.
Fixed billing for streams that end early.
Updated the status page.
Added Tor and I2P access to the full website and API.
Added an I2P address.
Updated the UI.
Updated the documentation.
Two-Factor Verification
Fixed the 2FA code window so it does not close by accident.
SilentAPI