If you're reading a foreign-language article and it's taking you four or five times longer than the equivalent piece in your native language, that's not a sign something is wrong with you or your method. It's the normal, expected state for most of the learning journey, and the frustration it causes is usually about mismatched expectations, not actual lack of progress.
There isn't a single "correct" reading speed to aim for, because native reading speed itself varies a lot by person and by text difficulty. What's useful instead is understanding roughly how speed tends to move relative to comprehension as proficiency grows, so you can tell the difference between normal early-stage slowness and an actual plateau.
Why speed lags comprehension, not the other way around
Comprehension and speed are not the same skill, even though it feels like one thing when you're reading in your native language. Comprehension is whether you can extract the right meaning at all. Speed is whether you can do it without consciously working through the sentence structure. At lower proficiency levels, you can often eventually get correct comprehension, given enough time, but the getting-there part is slow because you're actively parsing grammar rather than recognizing it.
This is why a learner can technically "understand" a paragraph and still take three times as long as a native speaker to do it. The comprehension is real. The automaticity that produces speed hasn't caught up yet, and it's a separate thing that develops through repeated exposure, not through understanding harder material.
What tends to happen at each rough stage
Early on, most of your reading time goes into word-level decoding: recognizing a word, retrieving its meaning, sometimes doing this more than once for the same word in one paragraph because it hasn't stuck yet. Reading here is genuinely slow and effortful, and pausing frequently is normal, not a sign you're doing it wrong.
In the middle stretch, word-level decoding gets faster for common vocabulary, but sentence-level structure, how clauses connect, where a subordinate clause ends, still requires conscious work for anything beyond simple sentences. This is usually where speed feels like it's improved a little but comprehension of longer or more complex sentences still lags. It's an uneven, not a smooth, improvement.
Later on, both word recognition and structural parsing start happening without deliberate effort for familiar sentence types, and speed starts approaching something closer to your native-language reading pace, though usually still somewhat slower, especially on unfamiliar topics or formal registers you haven't read much of.
The actual lever that moves speed: repeated exposure, not effort
A common mistake is trying to read faster by pushing through consciously, skimming harder, forcing the eyes to move quicker. This doesn't build the underlying automaticity; it just produces worse comprehension at the same slow underlying processing speed, because you're skipping the recognition step rather than speeding it up.
What actually moves the number is volume of exposure to real sentence patterns over time, the same mechanism that builds vocabulary recognition. Speed catches up as a side effect of enough reading, not as a target you train directly. This is part of why a daily reading habit, even short sessions, tends to outperform occasional long, effortful sessions for building actual reading speed: frequency of exposure matters more than total minutes in any single sitting.
A rough way to check your own trajectory
Rather than comparing your speed to a fixed benchmark, time yourself on a comparable piece of text, similar length, similar difficulty, once every few weeks. What you're looking for isn't a specific number, it's whether the trend is moving in the right direction over a month or two. A modest, steady drop in time for equivalent difficulty is the actual signal that the underlying skill is developing, even while any single session still feels slow in the moment.
Why comparing yourself to native reading speed backfires
Native reading speed in your own language took your entire childhood and years of schooling to build, and it applies to material you've had a lifetime of exposure to. Measuring your six-month-old foreign-language reading against that benchmark isn't just discouraging, it's comparing two fundamentally different amounts of accumulated exposure. A more useful comparison is your own speed a month ago on similarly difficult material, because that comparison actually reflects the thing you're trying to track: whether the specific skill is developing, not whether you've somehow failed to match decades of native exposure in a fraction of the time.
Topic familiarity affects speed as much as proficiency does
It's worth separating out how much of a slow reading session is genuinely about language proficiency and how much is about unfamiliarity with the topic itself. A learner who follows finance closely will read a financial news piece faster than an unrelated culture piece at the same objective language difficulty, simply because they already have the conceptual scaffolding for financial reporting in any language. When judging your own reading speed trend, comparing sessions on similar topics gives a cleaner signal than comparing a finance article one week to an unfamiliar science piece the next, where an apparent slowdown might reflect the topic switch rather than any real change in your language ability.