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/ 2026-10-29T12:00:00Z https://ai-fieldnotes.com/model-drops/#2026-10-29-canva-canva-brush 2026-10-29T12:00:00Z

Credit: Canva. Canva Research's vision-language model for design edits, from one line of copy to every slide in a deck. Live in Canva AI for Canva Pro subscribers. For film: Changes pitch decks and title cards on request, keeping what you didn't ask to change.

Canva https://ai-fieldnotes.com/model-drops/#2026-10-29-elevenlabs-sfx-v3 2026-10-29T12:00:00Z

Credit: ElevenLabs. New sound effects model in ElevenCreative and ElevenAPI that follows prompts describing several sounds at once. For film: One prompt can build a layered scene: crowd, ball, then boos, with fewer artefacts.

ElevenLabs https://ai-fieldnotes.com/model-drops/#2026-10-29-mistral-ai-mistral-large-4 2026-10-29T12:00:00Z

Credit: Mistral AI. Open weights for Mistral's 1-trillion-parameter multimodal model (text and image in, text out), on Hugging Face and the Mistral API.

Mistral AI https://ai-fieldnotes.com/model-drops/#2026-10-29-pika-pika-h3-flex 2026-10-29T12:00:00Z

Credit: Pika. Pika's post-trained MiniMax H3 video model with synced stereo audio, up to 12 image, video and audio references, and open weights. For film: Five separate audio stems through the API, so dialogue and score mix separately.

Pika https://ai-fieldnotes.com/model-drops/#2026-10-28-google-gemini-4-carbon 2026-10-28T12:00:00Z

Credit: Google. Google's new frontier Gemini model for coding, cybersecurity and long tasks; access starts with selected cyber defence partners.

Google https://ai-fieldnotes.com/model-drops/#2026-10-28-sensetime-sensenova-6-8-code2video 2026-10-28T12:00:00Z

Credit: SenseTime. Open 27B model that makes video by writing and fixing the code that draws each frame, aimed at motion graphics and explainers. For film: Editable motion graphics and explainer clips as code, not photoreal footage.

SenseTime https://ai-fieldnotes.com/model-drops/#2026-10-27-liquid-ai-lfm3 2026-10-27T12:00:00Z

Credit: Liquid AI. Open-weight family of small 2B-class models on one backbone: text, vision, audio, omni and decision models that run on phones and laptops. For film: LFM3-Audio-2B transcribes, voices and translates speech in 18 languages on-device.

Liquid AI https://ai-fieldnotes.com/model-drops/#2026-10-26-openai-gpt-6-1-terra 2026-10-26T12:00:00Z

Credit: OpenAI. Fast, cheaper GPT-6.1 model for visual reasoning, computer use and writing, with a 1.05M-token context; in ChatGPT Work, Codex and the API. For film: Half the API price of GPT-5.6 Terra; tested on time-coded shot logs and video editing.

OpenAI https://ai-fieldnotes.com/model-drops/#2026-10-24-moonshot-ai-kimi-k3-1-pro 2026-10-24T12:00:00Z

Credit: Moonshot AI. First model in the K3.1 series, with a 1M-token context and a faster Highspeed variant; live in Kimi and the Kimi API, full weights promised by 4 Nov. For film: Holds very long scripts and documents in one go; Moonshot cites gains in 3D creation.

Moonshot AI https://ai-fieldnotes.com/model-drops/#2026-10-22-canva-canva-craft 2026-10-22T12:00:00Z

Credit: Canva. Canva's first in-house image model: plans the layout first, then renders up to 2K across more than 30 aspect ratios. In Canva and the public API. For film: Planned layouts with legible text, useful for posters, title cards and pitch-deck stills.

Canva https://ai-fieldnotes.com/model-drops/#2026-10-21-tencent-hy-nomas 2026-10-21T12:00:00Z

Credit: Tencent. Open-weight interactive video world model: steer each new segment with text and camera moves as it generates, 720 x 1280 at 24 fps. For film: Change scenes, actions and camera mid-generation; an INT8 setup runs on one RTX 4090.

Tencent https://ai-fieldnotes.com/model-drops/#2026-10-20-bria-ai-fibo-2-0 2026-10-20T12:00:00Z

Credit: Bria AI. Open-weights image model that generates and edits in one, as Dev plus a 4-step Edit Turbo; non-commercial licence. For film: Instruction edits keep identity, layout and lighting; trained only on licensed data.

Bria AI https://ai-fieldnotes.com/model-drops/#2026-10-20-elevenlabs-eleven-realtime-v1 2026-10-20T12:00:00Z

Credit: ElevenLabs. New speech-to-speech models for voice agents: v1 Mini for natural chat at low cost, v1 Max for complex tasks. For film: Mini is pitched for interactive characters, in 90+ languages with any library voice.

ElevenLabs https://ai-fieldnotes.com/model-drops/#2026-10-20-mistral-ai-voxtral-3 2026-10-20T12:00:00Z

Credit: Mistral AI. New Voxtral speech family: Voxtral Small Transcribe 3, Voxtral Mini TTS 3 and a real-time voice agent SDK. For film: Voices shift emotion and pace with context; Pro Voice adapts from 30 minutes of audio.

Mistral AI https://ai-fieldnotes.com/model-drops/#2026-10-19-fal-and-black-forest-labs-flux-3-max 2026-10-19T12:00:00Z

Credit: fal and Black Forest Labs. Video model post-trained by fal Research on FLUX 3, for text and image to video at 720p. For film: A 10-second clip in about 12 seconds, so you can try many takes fast.

fal and Black Forest Labs 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. Open to everyone in the Kling app since 15 Oct: 3 to 30 second clips at 720p and 1080p, and Omni Reference with up to 15 references and 10 keyframes. 4K is listed but Kling's own 4.0 guide still marks it coming soon, and so is 10-bit HDR. Kling made the 4.0 API fully available on 28 Oct, at 720p and 1080p. For film: Long single takes with keyframe control, now in the app and on the API.

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/#2026-10-14-xai-grok-voice-3-0 2026-10-14T12:00:00Z

Credit: xAI. xAI's new speech-to-speech model, live in the API at the same price as Grok Voice 2.0. For film: Replies in under a second and handles pauses, turn-taking and interruptions naturally.

xAI 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 https://ai-fieldnotes.com/model-drops/#2026-10-06-google-deepmind-nano-banana-2-1 2026-10-06T12:00:00Z

Credit: Google DeepMind. An upgrade to the Gemini image line for visual design, mask-based editing and subject consistency, listed for Gemini, AI Studio, the Gemini API and Google Flow. For film: Mask edits and steadier subjects for boards and key art beside Flow.

Google DeepMind https://ai-fieldnotes.com/model-drops/#2026-09-30-utopai-studios-utopai-x 2026-09-30T12:00:00Z

Credit: Utopai Studios. Utopai Studios' text to video model with audio, built inside a working film and TV studio, launched with the new PAI production workspace that ties script, references and takes to each shot. For film: Takes stay tied to the shot, so you compare versions, not rebuild briefs.

Utopai Studios https://ai-fieldnotes.com/model-drops/#2026-09-28-elevenlabs-eleven-v4 2026-09-28T12:00:00Z

Credit: ElevenLabs. ElevenLabs' new text to speech model, plus a low-latency v4 Turbo, with more emotional range, stronger voice cloning and 90+ languages. Live in ElevenCreative, ElevenAgents and the API. For film: More emotional range for temp dialogue and voice-over drafts.

ElevenLabs https://ai-fieldnotes.com/model-drops/#2026-09-09-suno-suno-v6 2026-09-09T12:00:00Z

Credit: Suno. Suno's v6 music models, with v6-wild and v6-mini, adding more control over editing, mashups, sampling and lyric updates, and taking text, audio, image and video inputs. For film: Video and image inputs, plus finer edits, for temp score drafts.

Suno