How to Make Videos With Claude Code Using Fable 5: A Step-by-Step Guide

To make videos in Claude Code using Fable 5, you install a video generation skill — this guide uses Pexo (pexoai-agent) — select Claude Fable 5 as the model your agent runs on, then describe the video you want in plain language; Fable 5 interprets the request and drives the skill, and Pexo generates the finished result. One honest clarification up front: Fable 5 does not render video. Fable 5 is Anthropic's latest language model (id claude-fable-5), redeployed globally on July 1, 2026 across Claude.ai, Claude Code, and Claude Cowork; it is the reasoning and agent brain that reads your intent, plans the shots, and calls the Pexo skill. Pexo is the piece that actually produces the video, routing each shot across models like Seedance 2.0, Kling 3.0, Veo 3.1, and Sora 2. This guide walks the full workflow end to end: install the skill, set Fable 5, describe the video, let the agent generate, then review and export. If you first want the different ways Claude Code can make video, see can Claude Code make videos.

The Honest Division of Labor: Fable 5 Plans, Pexo Renders

The single most common misconception is that Fable 5 generates video. It does not. Claude Fable 5 is a language model, state-of-the-art on software-engineering and knowledge-work benchmarks, that runs the agent loop inside Claude Code — reading your request, decomposing it into a shot plan, and invoking tools. Pexo is the tool that turns that plan into pixels. The clean mental model: Fable 5 is the director who understands the brief; the Pexo pexoai-agent skill is the production crew that shoots, scores, and edits.

Because Fable 5 is a stronger reasoner, it tends to decompose a vague brief ("a punchy launch teaser") into a cleaner shot list before handing it to Pexo. That is the value Fable 5 adds — better interpretation of intent — not video rendering.

What You Need

Three things, and you can be running in a few minutes:

Claude Code does not generate video on its own, and neither does Fable 5 — the skill is what adds the video capability. Once it is installed and Fable 5 is set as the model, everything else happens inside the conversation. For the full field of options, see the best video generation skills for Claude Code.

Step 1: Select Claude Fable 5 as Your Model

Set your agent to run on Fable 5. Claude Fable 5 became available again on July 1, 2026 across Claude.ai, Claude Code, and Claude Cowork after a brief export-control pause, and through July 7, 2026 it is included for up to 50% of weekly usage limits on Pro, Max, Team, and select Enterprise plans — after that it moves to metered usage credits. In Claude Code, switch the active model:

# Inside Claude Code, choose the model your agent runs on
> /model
# Select "claude-fable-5" from the list

Setting Fable 5 changes only the reasoning brain, not the video engine — a more capable planner in front of the same Pexo rendering pipeline. If you are outside the July 7 window or on a plan without Fable 5, the identical workflow runs on any model; Fable 5 mainly improves how well the agent interprets a loose brief.

Step 2: Install the Pexo Skill

Add the Pexo skill (pexoai-agent) to your agent. Sign in at pexo.ai, copy your API key, and add the skill to your skills directory, then confirm the agent can see it:

# Inside Claude Code, list installed skills
> /skills
# You should see "pexoai-agent" in the list

The Pexo skill ships helper scripts plus a SKILL.md that tells the agent how to behave — including a rule to pass your request through faithfully rather than rewriting it. Once the skill is connected and your API key is set, the agent is ready to generate. Running the included diagnostic confirms the config, dependencies, and key are all valid before you start.

Step 3: Describe the Video You Want

This is the whole interface: tell the agent what you want in plain English, and Fable 5 interprets it into a shot plan. You do not pick a model or write a technical prompt. Useful things to specify:

• What it's about — "a launch teaser for our new running shoe"
• Length and shots — "15 seconds, three shots"
• Mood — "cinematic and premium," or "fast-paced for TikTok"
• Music — "ambient electronic," or let the agent choose
• Aspect ratio — "9:16 for Reels," "16:9 for YouTube"

A complete first request looks like this:

> Make a 15-second cinematic launch teaser for this running shoe —
  three shots, a slow orbit on the first, premium feel, ambient music. 9:16.

That is enough. Fable 5 reads the brief, decomposes it into a shot list, and hands it to the Pexo skill.

Step 4: Let the Agent Generate

Once you send the request, Fable 5 dispatches it to Pexo, which runs the full production: it writes a shot script, auto-selects the best model for each shot from 10+ options (a product close-up might route to Kling 3.0, a motion scene to Seedance 2.0, a cinematic wide to Veo 3.1, a narrative beat to Sora 2), generates each shot, adds transitions, composes a three-layer soundtrack (voiceover, music, and Foley sound effects), and mixes the audio.

A 15-second, three-shot video takes roughly 8–10 minutes end to end — far faster than choosing models, writing per-model prompts, and assembling clips by hand. While it runs, the agent polls for progress and reports back; you do not need to do anything until it returns the finished file. Fable 5's role here is orchestration and progress reporting, not rendering.

Step 5: Review and Iterate

When the video comes back, you review it and ask for changes the same way you described it — in plain language. There is no timeline to edit:

> Make the second shot slower, and swap the music for something more upbeat.

Fable 5 passes the revision through to Pexo and returns an updated cut. Because the request stays conversational, you iterate by talking, not by re-rendering anything yourself. A stronger reasoner tends to resolve ambiguous change requests ("make it feel more energetic") into concrete edits more reliably.

Step 6: Export and Use It

The finished video comes back as a standard MP4, mastered and ready to post. Because the production is multi-format aware, you can ask for the same content in different aspect ratios — 9:16 for TikTok and Reels, 16:9 for YouTube, 1:1 for feed — without regenerating from scratch. Download it and publish.

Step Summary

Tips and Troubleshooting

A few habits produce noticeably better videos, and a few common snags have quick fixes:

Other Ways to Start: Image, URL, Script, Audio

Text is only one input. Pexo accepts five, so Fable 5 can start from whatever you already have:

Scaling Up: From One Video to a Pipeline

Once a single video works, the same skill scales to batch production — dozens of ad variants from one set of inputs, or a video per SKU — all inside the conversation, with Fable 5 orchestrating each run. For the full automation pattern (product data in, finished ads out, staged for review), see how to build an AI video ad pipeline with Claude Code.

Related reading

• Can Claude Code Make Videos? The Three Ways, Compared
• How to Make Videos With Claude Code: A Step-by-Step Guide
• Best Video Generation Skills for Claude Code Agents
• How to Turn Photos into AI Video with Claude Code: Image-to-Video Guide
• How to Build an AI Video Ad Pipeline with Claude Code

Resources