B1 to B2 is the stage where the bottleneck changes. By B1 you already have the grammatical machinery you need for most sentences you will ever read. What separates you from B2 is quantity: how much text you have processed, how abstract a subject can get before you lose the thread, and how many registers you can move between without stalling. This plan is the direct sequel to the A2 to B1 reading plan, and it is deliberately heavier.
What actually changes between B1 and B2#
The CEFR descriptors are unusually clear here. B1 is about coping: handling most situations that arise while traveling, following clear standard speech on familiar matters, producing connected text on subjects of personal interest. B2 is about handling complexity: understanding the main ideas of demanding text on both concrete and abstract topics, following technical discussion in your own field, and arguing a viewpoint with the trade-offs laid out.
The word doing the work is abstract. B1 texts describe things that happened. B2 texts argue about what things mean. That shift brings a different vocabulary layer with it: nominalizations, hedging language, discourse connectives, and the whole apparatus a writer uses to signal that one clause qualifies another. None of it is difficult grammar. All of it is unfamiliar until you have seen a few hundred pages of it. If you want the descriptor set in full, what B2 level means unpacks it skill by skill.
Why this jump takes longer than the last one#
Two forces stretch the timeline. The first is vocabulary coverage. Paul Nation's research on vocabulary size and text coverage suggests that knowing roughly the first 2,000 to 3,000 word families gets you around 90% of the running words in general text, while getting to the 98% that supports comfortable independent reading takes something closer to 8,000 or 9,000. Those figures are approximate and vary by corpus, but the shape is consistent: each additional thousand words buys less coverage than the thousand before it. B1 sits low on that curve, B2 sits high, and the words in between are individually rarer.
The second force is exposure frequency. A word that appears once every 30,000 running words will not stick from reading alone, because you will not meet it often enough for the spacing to do its job. Below B1, passive reading was enough. Above it, reading has to be paired with deliberate review or the words evaporate.
Practically, published guided-hour guidance tends to put B1 at roughly 400 hours and B2 at roughly 600 for a language close to your own. The implication holds even if the numbers are loose: the B1-to-B2 band is wider than A2-to-B1, often close to double.
The 12-week structure#
The plan runs in three four-week blocks. Each escalates one variable at a time so the text never becomes incomprehensible all at once.
| Block | Weeks | Daily reading | Target-language share | Text diet | Weekly saved words |
|---|---|---|---|---|---|
| Range | 1–4 | 25 min | 55–65% | Opinion columns and long features on subjects you already follow | 25–35 |
| Abstraction | 5–8 | 30 min | 65–75% | Essays, popular science, reviews, long interviews | 30–40 |
| Register | 9–12 | 30–40 min | 75–85%, plus one unmodified text weekly | Novel chapters, editorials, forum threads, formal correspondence | 30–40 |
The target-language share is the lever that keeps difficulty in the productive band while the content gets harder. Raising abstraction and raising density at the same time is how learners burn out in week three. Raise one, hold the other, then swap.
Week-by-week targets#
| Week | Focus | Reading target | Additional task |
|---|---|---|---|
| 1 | Baseline | 25 min daily, familiar topics | Count unknown words in one 100-word passage |
| 2 | Volume | 25 min daily, same domain | Add a second short text on rest days |
| 3 | First domain switch | 25 min, one unfamiliar subject | Note which words were domain-specific |
| 4 | Consolidation | 25 min, mixed | Reread a week-1 text and compare effort |
| 5 | Argument structure | 30 min, opinion writing | Mark every connective you did not know |
| 6 | Abstraction | 30 min, essays or popular science | Summarize one text in five sentences |
| 7 | Density | 30 min at a higher blend share | Save collocations, not only single words |
| 8 | Consolidation | 30 min, mixed | Second 100-word coverage check |
| 9 | Informal register | 30 min, forums and comment threads | Collect slang and contracted forms |
| 10 | Formal register | 30 min, editorials or official text | Collect formulaic phrases |
| 11 | Native text | 35 min, one unmodified chapter | Read without stopping, look up afterward |
| 12 | Review | 40 min across all registers | Rerun the coverage check and compare all three |
The three coverage checks in weeks 1, 8, and 12 are the plan's only real metric. Take any hundred consecutive words of native text at a fixed difficulty, count the ones you do not know, and record the number. If it fell, the plan worked, regardless of how the reading felt on any given day.
The register problem most learners skip#
B1 learners typically read one kind of text: whatever their course or app serves them. B2 assumes you can move between an editorial, a group chat, and a set of instructions without a restart. The vocabulary overlap between those three is smaller than it looks, and the gaps are exactly the sort of thing that makes an otherwise strong reader stall on a novel.
Weeks 9 and 10 exist for this. Deliberately read something informal and something formal in the same week, and treat the discomfort as the point. Reading news is a reliable formal-register anchor, and reading news in a foreign language covers how to pick outlets that are demanding without being impossible.
Making the reading yield vocabulary#
At this level, reading exposes words and review retains them. The workflow that works is unglamorous: save every word you stop for, review on a spaced schedule, and stop saving words you can already recognize in context. A standard SM-2 ladder introduces a word at short intervals and stretches them as you succeed, which is what counters Ebbinghaus's forgetting curve (Ebbinghaus, 1885).
Two adjustments matter specifically at B1 to B2. Save collocations rather than isolated words, because B2 errors are usually combination errors rather than meaning errors. And shift some review into recall direction, prompting from your own language toward the target, since recognition alone will not carry into speech.
This is the loop LingoBlend automates: tap a word inside the text, get its meaning, grammatical form, and base form, save it in one tap, and let the spaced-repetition games schedule it. Reading text with a share of its words already in the target language is the diglot weave technique Robbins Burling described in 1968; the how it works page shows the full cycle.
The half this plan does not cover#
Reading builds receptive ability. The B2 descriptors also require interacting with fluency and producing clear, detailed argument, which no amount of reading delivers on its own. Budget two or three short speaking or writing sessions a week, drawing on what you read. Summarizing a week-6 essay out loud beats any textbook prompt, because the vocabulary is already active.
An honest note on timelines#
Twelve weeks at an average of 30 minutes a day is around 42 hours of reading. Against a B1-to-B2 band that plausibly runs 150 to 250 hours, that is a down payment rather than a completion. What it reliably delivers is a lower unknown-word count, a wider range of comfortable text types, and a review backlog worth several hundred words. Whether it certifies as B2 depends on your speaking, your listening, and the exam. The trade-off between reading widely and reading closely is covered in intensive versus extensive reading, and the earlier stage lives in the A2 to B1 plan.