Sep. 1–7 interview digest: timed coding screens, Google offer risk, and an Apple $320K package (2026)

The seven-day run-up to Sep. 7, 2026 produced clearer preparation signals than pass-rate statistics. A Reddit author described a rigid speedcoding test with two medium problems allotted 15 minutes each and three easy problems allotted 5–10 minutes each. A Google L4 candidate was still waiting for a decision after an onsite and a recruiter change. An OpenAI L4 candidate reached the final round and received a rejection. An Apple hardware candidate reported two parallel interview loops and a $320K TC figure after seven years of experience. 1 2 3 4

Blind displays relative ages rather than absolute native publication dates on the selected Google, OpenAI, Apple, Microsoft, and Netflix pages. The pages showed 4d, 5d, 6d, or 7d when retrieved, so those markers remain in the report. Reddit supplied absolute creation times for the Sep. 6 posts used here. Google, Meta, Amazon, and Stripe salary pages were updated Sep. 7, 2026. Glassdoor produced no usable current-window detail page through the fresh site-restricted search, and Levels.fyi supplied aggregate salary pages rather than first-person success stories. Comparable report-level votes and a company-level pass-rate denominator were unavailable, so the article uses recent, specific reports and labels each evidence boundary.

The week in one read

Four details change what a candidate should do before the next call.

Timed tests are becoming their own format problem. The Reddit author had already passed a phone screen before receiving an automated third-party test. The test reserved about 15 minutes for each of two medium-difficulty problems and 5–10 minutes for each of three easier problems. The author received a rejection after doing “decently well, but not perfect.” The post does not identify the employer or publish the questions. 1
The comment debate focused on signal quality and employer filtering. One commenter argued that the test selects for LeetCode speed rather than job performance. Another commenter, who described prior technical-interview work, said teams may use automated screens because thousands of applications arrive quickly. The comment tree returned 32 comments, with 24 more beyond the visible batch, so the discussion sample is incomplete. 1

“There were 2 of what I'd describe as Leetcode mediums that they allotted about 15 minutes for each.” 1

Google feedback can remain unresolved after a positive-looking loop. An L4 SWE-SRE candidate described a phone screen as Lean Hire, a hard onsite coding round as Lean Hire/Hire after a hint, and a second onsite coding round as Hire/Strong Hire. More than two weeks after the onsite, a recruiter change left the candidate waiting for feedback. Blind showed 6d and a generated summary based on 10 comments dated Sep. 4; the comments are discussion metadata, not a hiring denominator. 2
Final-round success and interview quality remain separate questions. An OpenAI L4 SWE candidate reported reaching the final round and receiving a rejection, while still having an active Anthropic process. The post gives no round-by-round question text, score, or compensation number. Its 49-comment generated summary contains supportive and skeptical reactions, but those reactions cannot establish an OpenAI pass rate. 3
Two teams can evaluate the same Apple candidate at once. A candidate with seven years of experience reported completing two Apple hardware loops, receiving positive feedback from both recruiters, and waiting for the second team's offer decision. The post lists $320K TC and asks how Apple handles team choice, leveling, and pay differences. Blind showed 7d; the native calendar date was unavailable. 4

Evidence at a glance

This week's evidence set

Current-window detail pages and the fields they expose.

First-person detail pages used	9
Concrete prompt or format signals	6
Fresh aggregate salary pages	4
Company pass-rate denominators	0

Cargando tarjeta de estadísticas…

The nine first-person pages comprise seven Blind posts and two Reddit posts. The Reddit post list also surfaced a Capital One Power Day advice request and a lead-software-engineer success post, but neither supplied a completed FAANG or top-SaaS interview loop with question-level evidence. The current Reddit interview thread itself was a moderator shell with zero comments at retrieval.

Report cards

Company or scope	Freshness shown	Round, task, or process	Reported outcome	Engagement and evidence boundary
Google	`6d` on Blind; native calendar date unavailable	L4 SWE-SRE phone screen plus two onsite coding rounds; recruiter changed while feedback remained pending.	Decision pending in the post.	Blind generated a summary from 10 comments dated Sep. 4. The summary supplies community reaction, while the author's round labels supply the interview facts. 2
OpenAI	`6d` on Blind; native calendar date unavailable	L4 SWE process; candidate reached the final round.	Rejected after the final round; Anthropic process remained active.	Blind generated a summary from 49 comments dated Sep. 4. The post gives no question text or score. 3
Apple	`7d` on Blind; native calendar date unavailable	Two hardware interview loops with separate teams; both recruiters gave positive feedback.	First team was moving toward an offer; second decision pending.	The author reports seven years of experience and $320K TC. Blind generated a 10-comment summary dated Sep. 1. 4
Microsoft	`4d` on Blind; native calendar date unavailable	L63 offer discussion after current Apple ICT3 compensation of about $250K TC.	Offer details were still being evaluated in the post.	Blind generated a summary from 41 comments dated Sep. 6. The summary mentions possible $250K–$270K outcomes, but the post itself does not publish a signed breakdown. 5
Netflix	`7d` on Blind; native calendar date unavailable	Discussion about design roles remaining posted or being reposted for long periods.	No individual interview outcome.	Blind generated a summary from 23 comments dated Sep. 6. The comments split between “ghost posting” suspicion and reports that real roles receive heavy application volume. 6
Unnamed employer	Posted Sep. 6 in the display timezone	Automated third-party coding test after a phone screen; two medium problems at about 15 minutes each and three easy problems at 5–10 minutes each.	Rejected after an imperfect result.	Reddit showed 40 points, a 90% upvote ratio, and 32 comments. The comment route returned only part of the tree. 1
Unnamed employer	Posted Sep. 6 in the display timezone	A team lead described slow decisions, interviewer PTO, late pipeline entry, and a final-round competition among three strong finalists.	The author's company hired the candidate who handled unfamiliar domains and ambiguity well.	Reddit showed 15 points and 11 comments; the comments route returned part of the tree. The author identified as a team lead under the hiring manager, not the final decision-maker. 7

The Reddit speedcoding thread is the only selected item with a comparable post score in this issue. Blind exposes relative age and generated comment counts, while Glassdoor did not expose report-level engagement. Those fields cannot be sorted into one cross-platform “top-voted” list.

The question bank you can practice now

1. Speedcoding test: five questions under fixed clocks

The test had two medium problems at roughly 15 minutes each and three easy problems at 5–10 minutes each. The author did not disclose the problem statements, so the useful preparation target is the format: read quickly, choose a tractable path, code without a long design phase, and leave time for a basic test. 1

Practice one 15-minute medium and one 7-minute easy problem with a visible countdown. At the end of each attempt, record the first point where time was lost: reading, algorithm choice, syntax, debugging, or test coverage. The format rewards a different balance from a 45-minute live coding conversation.

Cargando tarjeta de contenido…

2. Google L4 SWE-SRE: what happens after the coding rounds?

The candidate described a recruiter change and more than two weeks of waiting after an onsite. The reported round labels were Lean Hire, Hire, and Strong Hire, yet the post still had no decision. 2

Ask the recruiter four concrete questions: Has the hiring committee reviewed the packet? Is team matching part of this decision? Which feedback is still outstanding? What date should the candidate use for the next follow-up? These questions turn an open-ended wait into four answerable process fields.

3. OpenAI final round: how do you recover from a rejection?

The OpenAI candidate reached the final round and received a rejection without publishing the question set or the failure point. The active Anthropic process gives the candidate another process, but the post does not show whether the two companies use the same loop. 3

Prepare a post-interview review with one row per round: prompt type, first approach, correction received, final answer, and the evidence that remained weak. Ask for recruiter feedback where the company provides it. Keep “reached final round” separate from “passed final round”; they answer different questions.

4. Senior loop: how is ambiguity judged?

The Reddit team lead wrote that three finalists were strong and that the person hired handled unfamiliar domains and uncertain situations well. The post describes one employer's selection decision, not a universal Staff-level rubric. 7

“The final round is a competition; you're up against others and only the best one wins.” 7

For a senior interview, prepare one project story in which the requirements were incomplete. State what you knew, what you tested, which decision you made before certainty arrived, and how the decision changed after new evidence. That story gives the interviewer observable judgment instead of a generic claim about “handling ambiguity.”

Outcomes and the pass-rate read

The current pages describe four different endpoints: a Google candidate waiting for feedback after two coding rounds, an OpenAI candidate rejected after the final round, an Apple candidate moving through two parallel teams, and a speedcoding-test candidate rejected after an imperfect result. The endpoints cannot be combined into a pass rate because the companies, levels, stages, and denominators differ.

The engagement fields also measure different things. Reddit supplies post scores and upvote ratios. Blind supplies relative-age markers and generated comment counts. A 49-comment OpenAI discussion and a 10-comment Google discussion show attention on the platform; neither tells readers how many candidates interviewed or passed.

The best current outcome signal is qualitative: a positive onsite evaluation can still leave the decision open, a final-round appearance can still end in rejection, and two positive Apple recruiter signals can still leave the candidate waiting for team-specific terms. The practical preparation response is to ask for the stage owner, the remaining decision, and the date of the next update.

Compensation: one current disclosure, then the benchmark pages

First-person figures from this week

Company	Level and location	Base	Equity	Bonus	Sign-on	TC and status
Apple	Hardware role; location undisclosed	Not disclosed	Not disclosed	Not disclosed	Not disclosed	$320K TC reported by a candidate with 7 YOE after two parallel Apple loops; offer terms were still being worked through. 4
Microsoft	L63; current Apple ICT3 comparison	Not disclosed	Not disclosed	Not disclosed	Not disclosed	The post asks what compensation to expect and says the current Apple package is about $250K TC. The post does not publish a signed Microsoft offer breakdown. 5

The Microsoft discussion summary mentions commenters reporting possible L63 packages around $250K–$270K, with base salaries around $175K–$200K and varying stock and sign-on amounts. Those figures come from an unverified generated summary rather than the author's offer card, so they stay outside the first-person table. 5

Fresh aggregate annual TC anchors

Levels.fyi updated the four pages below on Sep. 7, 2026. The tables report aggregate US software-engineer submissions. They are level anchors for negotiation, not terms from the Apple or Microsoft candidates above.

Company	Level	Total	Base	Stock	Bonus
Google	L4 / SWE III	$286K	$190K	$76.1K	$19.5K
Google	L5 / Senior SWE	$427K	$230K	$168K	$29.9K
Google	L6 / Staff SWE	$640K	$272K	$315K	$53.4K
Meta	E4	$296K	$191K	$86.3K	$18.8K
Meta	E5	$432K	$227K	$179K	$25.4K
Meta	E6	$674K	$269K	$354K	$50.9K
Amazon	L5 / SDE II	$276K	$178K	$92.3K	$5.2K
Amazon	L6 / SDE III	$381K	$212K	$169K	$0
Amazon	L7 / Principal SDE	$714K	$259K	$455K	$0
Stripe	L2	$290K	$188K	$85.8K	$15.9K
Stripe	L3	$458K	$228K	$199K	$31K
Stripe	L4 / Staff Engineer	$662K	$274K	$330K	$58.4K

Google rows: 8 Meta rows: 9 Amazon rows: 10 Stripe rows: 11

Cargando gráfico…

The median figures in the chart come from the same four updated pages: Google $295K, Meta $420K, Amazon $265K, and Stripe $362K. The levels.fyi pages also show that stock timing differs by company and grant. Google lists several four-year examples, Meta lists four-year quarterly or annual patterns, Amazon shows backloaded schedules in its page context, and Stripe lists several vesting examples. Ask for the written schedule before comparing headline TC. 8 9 10 11

Company read-through

Company	Current-week signal	What a candidate can use	Evidence boundary
Google	Positive-looking coding labels were followed by more than two weeks of waiting and a recruiter change. 2	Ask whether feedback, hiring committee review, or team matching remains open.	One L4 SWE-SRE account; no company pass rate or standard timeline.
OpenAI	An L4 SWE candidate reached the final round and was rejected. 3	Review each round separately and request feedback before the next frontier-lab loop.	One outcome; the question set and failure point are undisclosed.
Apple	Two hardware teams gave positive recruiter feedback, with team-specific offer questions still open. 4	Ask whether both teams will disclose terms before the candidate chooses, and how level and pay are set.	One candidate's process; no general Apple policy is established.
Microsoft	A current Apple ICT3 candidate is comparing a possible L63 package with about $250K current TC. 5	Request base, equity, bonus, sign-on, vesting, and level before judging the move.	The post does not publish a signed Microsoft breakdown.
Netflix	The current discussion asks whether long-lived or reposted design roles are genuine; comments split between skepticism and reports of heavy applicant volume. 6	Ask whether the role has approved headcount, an active hiring manager, and a current interview slate.	One discussion; no verified hiring-status change.
Meta / Stripe / Amazon	Fresh aggregate salary pages provide level anchors; current-window first-person interview reports did not provide question-level detail for these companies. 9 10 11	Use the benchmark rows to frame a written-offer discussion, then compare vesting and cash timing.	Aggregate salary evidence; the selected current-window Blind prep post for Meta and Stripe was from Mar. 21, 2025 and was excluded from this issue.

The week links process uncertainty to preparation format. The speedcoding post makes time allocation the immediate problem. The Google and Apple posts make stage ownership and team choice the immediate problem. The OpenAI post makes post-final-round review the immediate problem. The compensation pages make written grant terms the immediate problem.

Before your next interview or negotiation

Run one timed drill. Use a 15-minute medium and a 7-minute easy problem. Review where the clock went before adding more questions.
Request the round map. Ask for format, duration, interviewer focus, coding environment, feedback timing, and the owner of the next decision.
Write an ambiguity story. Explain one project where requirements were incomplete. Include the first decision, the evidence behind it, and the later correction.
Separate intermediate outcomes. “Passed the phone screen,” “reached the final round,” “received positive feedback,” and “received an offer” belong in different notes.
Ask for the offer ledger. Record base, target bonus, equity grant, vesting schedule, sign-on payment, location, level, and refresh policy.
Compare year one with recurring value. A backloaded equity grant and a sign-on payment can make two packages with similar headline TC produce different cash flows.
Set a follow-up date. When a recruiter changes or a team decision remains open, ask for the next update date and the person responsible for it.

The current reports give candidates six immediate drills: fixed-clock coding, Google feedback follow-up, final-round postmortem, parallel-team questions, ambiguity stories, and component-level offer comparison. They also show the boundary around the evidence: recent community reports can sharpen preparation, while the current pages cannot support a cross-platform vote ranking or a company pass rate.

Fuentes de referencia

1
Are Automatic Speedcoding Tests Normal now?
reddit.com
2
Chance of Google offer
teamblind.com
3
Cried after failing my OpenAI interview
teamblind.com
4
Multiple Apple HW Interviews
teamblind.com
5
Recently got a Microsoft offer L63
teamblind.com
6
Are Netflix jobs just ghost postings?
teamblind.com
7
Why company takes SO long to get back to you
reddit.com
8
Google Software Engineer Salaries
levels.fyi
9
Meta Software Engineer Salaries
levels.fyi
10
Amazon Software Engineer Salaries
levels.fyi
11
Stripe Software Engineer Salaries
levels.fyi