Skip to content

Gemini

model family · in 25 sources (2025–2026)

Adil uses Gemini to write detailed animation prompts, ending his request with "Do not generate, just give me the prompt" so it does not animate the image itself. Since mid-2026 its video member, Gemini Omni Flash, runs on Higgsfield for editing clips through plain-language follow-ups; version 1.1 went live in August 2026.

Made by: Google · Includes: Gemini Omni Flash, Gemini Omni Flash 1.1, Gemini 3.1 Pro, Gemini 3.5 Flash, Gemini 3.0 Flash

Sources by year: 2025: 4 · 2026: 21

Threads

Also named in the threads: Nano Banana on Higgsfield

Often together

What comes up in the same parts of the videos; the number is how many parts.

Google (9) · prompt adherence (3) · Higgsfield Moodboards (3) · Starter plan (5) · OpenAI (5) · ChatGPT (7) · Seedance (6) · Nano Banana Pro (8) · Nano Banana 2 (4) · GPT Image (4)

What the videos say

The most important statements of each year; all statements.

2026

  • 2026-02-25 · Nano Banana 2: Pro-Level Al Image Generation with Gemini 3.1 Flash. Let's Peel It Back — announcement: Higgsfield: Google has introduced Nano Banana 2 (Gemini 3.1 Flash Image), its new image model, positioned between the original Nano Banana and the studio-grade Nano Banana Pro.
  • 2026-03-01 · AI Videos Look Bad? Here's Why — recommendation: Adil's prompt hack: bring the image to animate to Gemini, ChatGPT or another LLM, describe the animation simply and ask for a detailed prompt, then read and tweak it rather than copy-paste it blindly. — 00:06:52
  • 2026-06-09 (said on 2026-10-20, implied) · The 10 Best AI Video Models in 2026, Tested and Compared — recommendation: Higgsfield's blog calls Gemini Omni Flash 1.1 the best model for iterative editing through conversation: attach text, image, audio and video, then refine with follow-ups in one thread.
  • 2026-06-30 · Changelog: Gemini Omni Flash. One prompt in, a whole scene out — announcement: Higgsfield: Gemini Omni Flash arrives on Higgsfield in Preview (beta): 720p clips of 4, 6, 8 or 10 seconds, in 16:9 or 9:16, with native audio.
  • 2026-07-13 · How to Use Gemini Omni Flash for VFX and Video Editing (2026 Guide) — Higgsfield's blog says Gemini Omni Flash fixes VFX problems such as distortion under motion and broken physics through plain-language edits on an existing clip, without regenerating the whole shot.
    • After a Gemini Omni Flash generation, a change described in plain language is applied as a targeted edit, holding character consistency, physics, scene context and composition.
    • To replace a clip's background with Gemini Omni Flash, describe the new environment; the blog says the foreground subject, lighting relationship and physics hold, with no rotoscoping or masking.
    • Gemini Omni Flash takes up to 7 reference images per prompt plus text and video, and outputs video with synchronized audio in one pass, up to 10 seconds, landscape or portrait.
    • Higgsfield's blog says Kling 3.0 handles human subjects more accurately than Veo 3.1, Seedance 2.0 or Gemini Omni Flash: body movement, micro-expressions, skin tones and complex gestures.
    • Gemini Omni Flash costs about $0.15 per second, capped at 720p and 10 seconds, with lower peak quality than Veo 3.1; the blog places it in development and revision, not final delivery.
  • 2026-08-28 · Changelog: Gemini Omni 1.1 Flash. Now on Higgsfield — announcement: Higgsfield's changelog of 28 August 2026 says Gemini Omni 1.1 Flash, Google's multimodal model for video generation and editing, is live on Higgsfield.
  • 2026-10-22 · Gemini Omni Flash 1.1 4K Direct every frame — Higgsfield: Gemini Omni Flash 1.1 reads up to ten seconds of the preceding footage, not just the last second, and continues its motion, light and story in ten-second steps, holding one take as long as needed.
  • …and 52 statements more in 2026 — all statements of 2026

2025

Also named in: A $1,000,000 AD Using Just 2 tools · Full AI Workflow