Young Tries AI

Gemini Omni 1.2 Flash, Images 2.6 and ElevenLabs SFX v3 land for creators

The week in one breath: Video from Google, images from OpenAI, sound effects from ElevenLabs and story-driven ads from Higgsfield all arrived this week. Anthropic also released Claude Fable 5.2 for the hardest coding work. Across the week's daily issues, 37 of 73 items can be used today.

73 items · 37 usable today · 7 guides · 35 finds

  1. ElevenLabsSFX v3Sound Effects API

    Video and imageFor creators, developers

    ElevenLabs released SFX v3 in the Sound Effects API, and it follows prompts with multiple instructions more closely than SFX v2. It handles vocals and music better and is less likely to clip on high-frequency sounds such as gongs, chimes and kettle whistles. Creators can also use it in the Sound Effects tab, Flows and Studio in ElevenCreative, but looping still works only with the v2 model.

    Can you use it today?Yes

    Change model_id to eleven_text_to_sound_v3 in an existing sound-effects call, or use the Sound Effects tab in ElevenCreative, and compare your usual ad prompts against v2.

    Source: ElevenLabs Blog한국어로 읽기

  2. Google GeminiGemini Omni 1.2 Flash

    Video and imageFor creators, business and marketing, developers

    Gemini Omni 1.2 Flash is generally available in the Gemini API and Gemini for business, with video generation and editing up to 15 seconds, up from 10 seconds in Omni 1.1. Price is $0.07 per second for 720p, a 30% drop from Gemini Omni 1.1, and drafts can be made at 360p or 720p with the best ones upscaled to 1080p or 4K. Chat editing works on full 15-second clips, so a product, background, lighting or aspect ratio can change without regenerating.

    Can you use it today?Yes

    Call Gemini Omni 1.2 Flash through the Gemini API or Gemini for business to draft 15-second ad variants at 720p, then edit by chat and upscale the keepers.

    Source: Google DeepMind, blog.google한국어로 읽기

  3. OpenAIImages 2.6GPT-Image-2.6 FlareGPT-Image-2.6 Sunburst

    Video and imageFor creators, business and marketing, beginners

    Images 2.6 is available in ChatGPT and Codex across all subscription tiers, while GPT-Image-2.6 Sunburst and GPT-Image-2.6 Flare are in the API. OpenAI says it generates images 50% faster than before, and with quality set to low, Flare can produce images in as little as six seconds. Sunburst targets consistency across edits, Flare improves layout for posters and product imagery, and the API adds SVG exports and image markup to pinpoint edits.

    Can you use it today?Yes

    Use Images 2.6 in ChatGPT or call GPT-Image-2.6 Flare or Sunburst in the API for product stills, targeted edits, SVG export and GIF loops.

    Source: OpenAI News한국어로 읽기

  4. HiggsfieldAds StudioMicro Drama AdsHiggsfield MCP

    Video and imageFor business and marketing, creators

    Higgsfield Ads Studio now makes micro drama ads, story-driven video ads with characters, a conflict and the product at the turning point. Describe the ad or paste a script, pick a style, a product and an avatar, and an agent in the chat makes the video and new versions on request. It offers eight video styles and 44 story cards, and it runs on the web and in Claude and ChatGPT through Higgsfield MCP.

    Can you use it today?Yes

    Open Ads Studio on the web, or run /microdrama-ads through Higgsfield MCP in Claude or ChatGPT, then pick a story card, a brand-kit product and an avatar.

    Source: Higgsfield Blog한국어로 읽기

  5. ClaudeClaude Fable 5.2Claude Mythos 5.2Claude Opus 5.5

    ModelsFor developers, office workers

    Claude Fable 5.2 and Mythos 5.2 are the same underlying model; Fable is available to all users, while Mythos access is by application. They generate outputs up to 30% faster than the 5.1 versions and cost $8 per million input tokens and $40 per million output tokens. Anthropic recommends starting most coding and knowledge work on Opus 5.5 and moving to Fable 5.2 for the hardest projects, long unattended runs or major refactors.

    Can you use it today?Yes

    Select Claude Fable 5.2 in Claude or the API for heavy Claude Code refactors, keeping Opus 5.5 as the default and trying effort levels before paying for the bigger model.

    Source: Anthropic News, claude.dev, claude.dev한국어로 읽기

  6. Claude Codev2.1.317Agent SDK

    CodingFor developers

    Claude Code v2.1.317 adds onFailure set to block for Agent SDK hook callbacks, so a callback that fails blocks the action instead of letting it run. It also warns when a new cloud scheduled task in /schedule gets connectors you did not ask for, and asks before a web or shell tool call sends your account email address to the network. Among the fixes, claude -p now exits 1 instead of 0 when a full disk or quota loses its transcript.

    Can you use it today?Yes

    Update Claude Code to v2.1.317 and set onFailure: block on guard hook callbacks so a failing guard stops the action.

    Source: Claude Code releases한국어로 읽기

This week’s guide

Intermediate guide for creators and businesses: build a pack of animated chat stickers from a single character image with Animated Sticker Studio in about 20 minutes.

Finds of the week

New skills, plugins, MCP servers, repos and small products, described only from their own README or landing page.

The daily archive holds all 73 items the week's issues covered, plus seven step-by-step guides and 35 picked new tools.