New AI video, image, audio and language model releases, one dated line each, linked to the maker. Links only. https://ai-fieldnotes.com/model-drops/ 2026-10-19T12:00:00Z https://ai-fieldnotes.com/model-drops/#2026-10-19-minimax-minimax-m3-1-flash 2026-10-19T12:00:00Z

Credit: MiniMax. New Flash language model for agents and coding; MiniMax says nearly three times faster online and 70% cheaper to use.

MiniMax https://ai-fieldnotes.com/model-drops/#2026-10-17-deepseek-deepseek-v4-1-flash-1017 2026-10-17T12:00:00Z

Credit: DeepSeek. V4.1 Flash updated to version 1017, now behind the deepseek-flash API name, with weights on Hugging Face.

DeepSeek https://ai-fieldnotes.com/model-drops/#2026-10-15-black-forest-labs-flux-3-image-klein 2026-10-15T12:00:00Z

Credit: Black Forest Labs. 5.6B open-weights image model for text-to-image, editing and multi-reference with up to 10 images, native 4K. For film: Runs locally on consumer GPUs, and bounding boxes place elements where you ask.

Black Forest Labs https://ai-fieldnotes.com/model-drops/#2026-10-15-bytedance-byteplus-omnihuman-2-0 2026-10-15T12:00:00Z

Credit: ByteDance (BytePlus). Audio-driven avatar video from one image plus speech, music or dialogue, now on the BytePlus API. For film: Lip movement and expression follow the audio, for talking and singing characters.

ByteDance (BytePlus) https://ai-fieldnotes.com/model-drops/#2026-10-15-elevenlabs-instant-soundtracks 2026-10-15T12:00:00Z

Credit: ElevenLabs. Upload a video and get an original score plus sound effects placed on the timeline, in ElevenCreative and the API. For film: Effects come as separate layers you can retime; the score can follow existing audio.

ElevenLabs https://ai-fieldnotes.com/model-drops/#2026-10-15-krea-krea-3 2026-10-15T12:00:00Z

Credit: Krea. Krea's new image editing model: most edits under 8 seconds at native 2K, up to 6 reference images per edit. For film: Keeps the same character across poses, locations and light for a consistent set.

Krea https://ai-fieldnotes.com/model-drops/#2026-10-15-kuaishou-kling-4-0 2026-10-15T12:00:00Z

Credit: Kuaishou. Kling 4.0 fully live on the Kling site: native 30-second takes, Omni Reference with up to 15 references and 10 keyframes. For film: One-pass long takes and keyframe control for continuous scenes; API access is promised soon.

Kuaishou https://ai-fieldnotes.com/model-drops/#2026-10-15-meta-muse-spark-open 2026-10-15T12:00:00Z

Credit: Meta. Open-weight multimodal model, 816B total and 42B active, 1M-token context, built for coding and agentic work.

Meta https://ai-fieldnotes.com/model-drops/#2026-10-14-heygen-heygen-voice 2026-10-14T12:00:00Z

Credit: HeyGen. Voice cloning that turns a recording, a script and a target image into speech shaped for the scene on screen. For film: Delivery and room acoustics match the shot, from studio to street to a moving car.

HeyGen https://ai-fieldnotes.com/model-drops/#2026-10-14-mirelo-mirelo-sfx-2-0 2026-10-14T12:00:00Z

Credit: Mirelo. Video-to-sound model that replaces a clip's audio with synced effects, foley and ambience while keeping dialogue. For film: Keeps dialogue while adding sound; Studio and API take clips up to 10 minutes.

Mirelo https://ai-fieldnotes.com/model-drops/#2026-10-13-fish-audio-drama-3 2026-10-13T12:00:00Z

Credit: Fish Audio. Fish Audio's most emotive text-to-speech model, live on fish.audio, the Fish Audio app and the API. For film: Follows direction: a whisper stays a whisper and a laugh lands where you put it.

Fish Audio https://ai-fieldnotes.com/model-drops/#2026-10-13-google-gemini-4-argon 2026-10-13T12:00:00Z

Credit: Google. Google's frontier model now generally available in the Gemini API, AI Studio and the Gemini app for Pro and Ultra.

Google https://ai-fieldnotes.com/model-drops/#2026-10-11-alibaba-wan-3-0-prime 2026-10-11T12:00:00Z

Credit: Alibaba. Faster Wan 3.0 video model in Model Studio: text, image, video and audio in; 480P to 1080P, up to 30 seconds. For film: Text-to-video, first-last frame and reference video in one model, generated faster.

Alibaba https://ai-fieldnotes.com/model-drops/#2026-10-07-anthropic-claude-haiku-5-5 2026-10-07T12:00:00Z

Credit: Anthropic. Anthropic's fastest small model yet, around 75% cheaper to run than Haiku 4.5 on average.

Anthropic https://ai-fieldnotes.com/model-drops/#2026-10-07-shengshu-vidu-q4-preview 2026-10-07T12:00:00Z

Credit: Shengshu. Preview of the next Vidu model: image and reference-to-video with native audio, 3 to 16 second clips, 2K/4K out. For film: Up to 15 image and 3 voice references, with multi-shot camera moves in one take.

Shengshu