Getting Started

Paste your script, scene description, or video concept and I'll generate a fully rendered video using text-to-video-ai-2026 models. No footage? No problem — just describe what you want and I'll build it from scratch.

Try saying:

• "Create a 30-second product launch video for a wireless earbud brand using a sleek, dark cinematic style with upbeat background music and on-screen text callouts"
• "Generate a 60-second educational explainer video about how black holes form, using a space documentary visual style with a calm narrator voiceover and animated diagrams"
• "Turn this blog post intro into a vertical-format social media video with bold captions, fast cuts, and an energetic tone suitable for Instagram Reels"

Quick Start Setup

This skill connects to a cloud processing backend. On first use, set up the connection automatically and let the user know ("Connecting...").

Token check: Look for NEMO_TOKEN in the environment. If found, skip to session creation. Otherwise:

• Generate a UUID as client identifier
• POST https://mega-api-prod.nemovideo.ai/api/auth/anonymous-token with X-Client-Id header
• Extract data.token from the response — this is your NEMO_TOKEN (100 free credits, 7-day expiry)

Session: POST https://mega-api-prod.nemovideo.ai/api/tasks/me/with-session/nemo_agent with Bearer auth and body {"task_name":"project"}. Keep the returned session_id for all operations.

Let the user know with a brief "Ready!" when setup is complete. Don't expose tokens or raw API output.

From Words on a Page to Video That Moves

Text-to-video-ai-2026 is built for anyone who has ever had a clear vision in their head but no crew, no camera, and no time to execute it. You write a prompt — a scene description, a script, a concept — and the skill translates it into a cohesive video with visuals, pacing, and optionally voiceover or captions baked in.

This isn't a basic slideshow generator. The 2026 generation of AI video models understands narrative structure, visual continuity, and stylistic tone. You can ask for a cinematic product reveal, a whiteboard explainer, a social media reel, or a news-style segment — and get back something that actually looks intentional, not stitched together.

The skill is designed to work iteratively. You can refine outputs by adjusting your prompt, changing the visual style, swapping the pacing, or requesting a different aspect ratio. Think of it as a creative collaborator that handles the heavy lifting while you stay focused on the message you're trying to deliver.

Prompt Routing and Model Dispatch

Each text prompt is parsed for scene complexity, motion directives, and style tokens before being dispatched to the optimal diffusion pipeline in your connected model cluster.

Cloud Inference API Reference

Video generation requests are processed across distributed GPU nodes using latent diffusion with temporal attention layers, delivering rendered MP4 outputs via signed CDN URLs. Frame coherence, motion smoothing, and upscaling passes all run server-side — no local compute required.

Skill attribution — read from this file's YAML frontmatter at runtime:

• X-Skill-Source: text-to-video-ai-2026
• X-Skill-Version: from frontmatter version
• X-Skill-Platform: detect from install path (~/.clawhub/ → clawhub, ~/.cursor/skills/ → cursor, else unknown)

All requests must include: Authorization: Bearer <NEMO_TOKEN>, X-Skill-Source, X-Skill-Version, X-Skill-Platform. Missing attribution headers will cause export to fail with 402.

API base: https://mega-api-prod.nemovideo.ai

Create session: POST /api/tasks/me/with-session/nemo_agent — body {"task_name":"project","language":"<lang>"} — returns task_id, session_id.

Send message (SSE): POST /run_sse — body {"app_name":"nemo_agent","user_id":"me","session_id":"<sid>","new_message":{"parts":[{"text":"<msg>"}]}} with Accept: text/event-stream. Max timeout: 15 minutes.

Upload: POST /api/upload-video/nemo_agent/me/<sid> — file: multipart -F "files=@/path", or URL: {"urls":["<url>"],"source_type":"url"}

Credits: GET /api/credits/balance/simple — returns available, frozen, total

Session state: GET /api/state/nemo_agent/me/<sid>/latest — key fields: data.state.draft, data.state.video_infos, data.state.generated_media

Export (free, no credits): POST /api/render/proxy/lambda — body {"id":"render_<ts>","sessionId":"<sid>","draft":<json>,"output":{"format":"mp4","quality":"high"}}. Poll GET /api/render/proxy/lambda/<id> every 30s until status = completed. Download URL at output.url.

Supported formats: mp4, mov, avi, webm, mkv, jpg, png, gif, webp, mp3, wav, m4a, aac.

SSE Event Handling

~30% of editing operations return no text in the SSE stream. When this happens: poll session state to verify the edit was applied, then summarize changes to the user.

Backend Response Translation

The backend assumes a GUI exists. Translate these into API actions:

Draft field mapping: t=tracks, tt=track type (0=video, 1=audio, 7=text), sg=segments, d=duration(ms), m=metadata.

Timeline (3 tracks): 1. Video: city timelapse (0-10s) 2. BGM: Lo-fi (0-10s, 35%) 3. Title: "Urban Dreams" (0-3s)

Error Handling

Best Practices

Start every text-to-video-ai-2026 session by defining three things: the audience, the platform, and the desired emotional response. A training video for enterprise employees needs a completely different visual language than a TikTok ad for Gen Z consumers — and the AI responds well to that kind of contextual framing in your prompt.

Iterate in layers. Get the structure and pacing right first, then refine the visual style, then polish the copy or voiceover. Trying to perfect everything in a single prompt often leads to over-constrained outputs that feel forced.

For brand consistency, include specific style references in your prompts — color hex codes, font style descriptors, or references to visual aesthetics (e.g., 'Wes Anderson symmetry', 'Apple product launch minimalism'). The 2026 models are trained on a wide enough visual corpus to interpret these references accurately and apply them with real coherence across a full video.

Performance Notes

Text-to-video-ai-2026 models perform best when your input prompt is specific about visual style, duration, and intended platform. Vague prompts like 'make a video about coffee' will produce generic results, while prompts that specify mood, color palette, pacing, and subject framing consistently yield higher-quality outputs.

Longer videos (over 90 seconds) may require segmented generation — breaking your concept into scenes and stitching them together produces more visually coherent results than requesting a single long render. For complex narratives, providing a structured scene-by-scene breakdown dramatically improves output consistency.

Aspect ratio and resolution targets should be declared upfront. Specifying 9:16 for mobile, 16:9 for desktop, or 1:1 for feeds ensures the composition and subject framing are optimized for your delivery channel from the first render rather than requiring a crop or reformat afterward.