The Arabic Input Roadmap.
Seven levels defined by what you can understand and do, with typical hour ranges and the full methodology behind every number. One register at a time: Levantine or MSA.
Our job, stated precisely
Arabic All The Time carries you to C1-level comprehension of the Arabic register you choose, Levantine or MSA, through comprehensible input, and gives you the foundation from which speaking emerges. Progress is measured in hours of input and in what you can understand. Never in lessons completed, words memorized, or streaks.
What you can do, level by level.
Hour bands are typical ranges for consistent learners, not promises. Real variation runs in both directions, and the bands widen as the levels rise because variation compounds.
1. Tuning In
0 to 75 hours
Arabic stops being an undifferentiated stream. You follow A1 videos fully through visuals and slow speech, and you hear where words begin and end. Greetings, numbers, and common objects become recognizable.
Self-check: “I can follow an A1 video without pausing and know what happened.”
2. First Words Stick
75 to 250 hours
Easy learner videos are comfortable. High-frequency words are recognized automatically across different videos and speakers, without trying.
Self-check: “Common words jump out at me before I think about them.”
3. Simple Stories
250 to 500 hours
You follow simple stories and descriptions on familiar topics made for learners, at more natural pacing. Word-family patterns such as kitab, katib, and maktab start feeling intuitive rather than learned.
Self-check: “I can follow a new simple story on a familiar topic the first time through.”
4. Everyday Topics
500 to 1,000 hours
You follow patient conversation and advanced learner content on everyday topics. Words and phrases begin surfacing on their own. For many learners, this is where speech starts emerging unprompted.
Self-check: “I can follow a patient conversation between two speakers about daily life.”
5. Bridging to Native
1,000 to 1,700 hours
All learner content is easy. Native content on familiar topics becomes accessible with effort. You follow the thread and most of the detail in your register.
Self-check: “I can watch native content on a topic I know and follow the thread.”
6. Native Content, Your Terms
1,700 to 2,500 hours
You follow most native media in your register on general topics without strain. If you have been speaking, conversation feels natural on familiar ground: imperfect and real.
Self-check: “I can put on a show or podcast in my register and mostly just enjoy it.”
7. Broad Comprehension
2,500+ hours
Comprehension in your register is broad, automatic, and durable, including unfamiliar topics, at roughly the C1 level of the CEFR listening descriptors. Refinement continues by living the language rather than counting hours.
Self-check: “Topic no longer decides whether I understand.”
Where the good part starts: the things most learners actually want, real conversations on familiar ground and native media you can enjoy, arrive across Levels 4 to 6. Level 7 is the completion of the platform's job, not the entry ticket to using Arabic.
Where these numbers come from.
Most timelines on the internet skip this section. We think it is the most important one on the page.
The anchor
The Dreaming Spanish roadmap, with levels at 50, 150, 300, 600, 1,000, and 1,500 hours for Spanish, is the largest body of observational milestone data for input-based acquisition. No comparable dataset exists yet for Arabic, so every Arabic hour estimate anywhere, including ours, is an extrapolation. The question is which extrapolation is best reasoned.
The difficulty gap
The U.S. Foreign Service Institute's categories are the best available measure of linguistic distance from English: Category I Spanish at roughly 600 to 750 class hours, and Category IV Arabic at roughly 2,200 class hours over 88 weeks. The target is four-skill professional proficiency, historically in MSA plus a spoken variety. The raw ratio is about 3x. These are class hours, and FSI students also study substantially outside class.
Vocabulary anchors
Paul Nation's coverage research suggests roughly 1,000 word families support simple learner content, roughly 3,000 support conversational breadth, and roughly 8,000 to 9,000 support broad native media. We cite these as anchors only, deliberately unmapped from specific hour counts, because learners cannot measure their own word counts.
Why we scale by roughly 1.6 to 1.8x, not 3x
The FSI Arabic figure prices in things this roadmap does not promise: professional four-skill proficiency, strong formal literacy, and command of two registers at once. Our roadmap targets comprehension-first progress in a single chosen register. Removing the costs we do not charge for, and keeping the genuine ones that remain, we scale the Spanish milestone bands by roughly 1.6 to 1.8x, with ranges that widen at higher levels. That is how 50 becomes 75, 600 becomes 1,000, and 1,500 becomes 2,500. This is a reasoned estimate, not a measured constant, and we say so.
What we know we do not know
Prior language experience, input intensity, content fit, and consistency all move the bands, and no published milestone data for Arabic comprehensible input exists yet. When we are uncertain, the roadmap says so instead of rounding the uncertainty away.
How the numbers get better
The AATT progress tracker measures real comprehensible input minutes, and optional level self-assessments are being added at band boundaries. We will publish the first calibration report comparing tracked hours to self-assessed levels, then revise these bands against real data. When the data disagrees with the estimate, the data wins.
Version history
v2.0 (July 2026). First public methodology. Bands revised upward from v1, which followed the Dreaming Spanish milestones too closely while describing them as difficulty-scaled. A member of our community pointed out the inconsistency, and they were right. Per-level word counts were removed. Changes to this roadmap are logged here.
One register at a time.
The roadmap describes one track: the Levantine track or the MSA track. If you split hours across both, each track advances according to the hours it receives, with meaningful but partial transfer between them. Shared core vocabulary and structure carry over. Sounds, function words, and register habits carry over less.
Adding a strong second register after finishing the first is substantially faster than starting over, but not free. As a working estimate, it takes roughly a third to a half of the original track, concentrated at the lower levels. Like everything on this page, that estimate is provisional and will be calibrated against learner data.
The map is honest. The hours are yours.
Start at Level 1 today, in Levantine or MSA, and let the tracker count what actually matters.
