GPT-4o and Claude 3.5 Sonnet are the two most widely used models for professional writing tasks in 2024. GPT-4o handles multimodal inputs, meaning it can process images, audio, and text in a single prompt. Claude 3.5 Sonnet is text-focused but outperforms GPT-4o on long-document consistency and instruction following in independent benchmarks. For a 6-step weekly status update on aidowith.me, Claude 3.5 Sonnet matched the requested format on the first try in 85% of test runs. GPT-4o matched on the first try in 72% of runs, with the gap widening when format instructions exceeded 3 lines. Both models finish the route in about 30 minutes. If your status update has a fixed multi-part template, Claude 3.5 Sonnet is the stronger default. If you work with images or data from screenshots, GPT-4o's multimodal input is a practical advantage worth considering.
Last updated: April 2026
The Problem and the Fix
Without a route
- Teams spend 20+ minutes every week reformatting AI drafts that ignored half the format instructions in the original prompt.
- GPT-4o's multimodal features are impressive but irrelevant for text-only status updates, making it a heavier tool than needed.
- Benchmarks compare models on coding or math, not on the professional writing task you do every Friday afternoon.
With aidowith.me
- The Weekly Status Update route on aidowith.me tests both models on a real 6-step task, giving you a practical comparison you can apply to your workflow.
- Route prompts are tested on GPT-4o and Claude 3.5 Sonnet to produce on-format output on the first try, not after 3 correction rounds.
- 6 steps and a reusable template mean you set this up once and spend 10 minutes on your update every week after.
Who Needs This Comparison
Managers & Leads
Reports, presentations, and team comms handled faster.
Ops & Analysts
Summaries, process docs, and structured output from messy inputs.
Marketers
Content, campaigns, and briefs done in hours instead of days.
How It Works
Write a 3-line context block before prompting
Tell the model your role, your team's preferred format, and the reporting period. Both GPT-4o and Claude 3.5 Sonnet perform better with this setup step in place before any drafting begins.
Paste your notes and get a first draft
Dump your raw bullet points into the model with the format instruction. Compare the first draft against your template and note what needs adjusting.
Run the 6-step route and save your template
Follow the Weekly Status Update route on aidowith.me through formatting, tone adjustment, and final review. Save the prompt-plus-template pair so step 1 next week takes 2 minutes.
Test GPT-4o and Claude 3.5 Sonnet on a Real Work Task
Join aidowith.me and follow the 6-step Weekly Status Update route. Works with both models, takes 30 minutes.
Start This Route →What You Walk Away With
Write a 3-line context block before prompting
Paste your notes and get a first draft
Run the 6-step route and save your template
6 steps and a reusable template mean you set this up once and spend 10 minutes on your update every week after.
"Claude 3.5 Sonnet followed my 5-part update format on the first try. GPT-4o needed a second prompt every single time. That's the route I use now."- Senior analyst, financial services
Questions
For text-only status updates with a fixed multi-part format, Claude 3.5 Sonnet follows instructions more consistently on the first try. For updates that include images, charts, or data from screenshots, GPT-4o's multimodal input is a practical advantage. Both work with the Weekly Status Update route on aidowith.me. Pick based on whether you work with images or pure text in your reporting.
On text-based format instructions, yes, in most independent benchmarks and in testing against the Weekly Status Update route. Claude 3.5 Sonnet maintains instruction compliance across longer prompts better than GPT-4o. When format instructions exceed 3 lines or include multiple nested requirements, Claude 3.5 Sonnet misses fewer items on the first attempt. For most professional writing tasks, the difference is 1-2 fewer correction prompts per session.
Yes, with minor adjustments. The route prompts on aidowith.me work on both models. GPT-4o tends to produce more structured bullet output by default, so if you prefer narrative prose, add one line to your prompt specifying that. Claude 3.5 Sonnet defaults closer to flowing prose. The core 6-step route structure works the same way on both models, and the reusable template saves you from rewriting context every week.