"How much does it actually pay?" is one of the hardest questions to answer in tech. Posted salary ranges are often misleading. Equity valuations shift with each funding round. Stock refreshers, signing bonuses, and performance multipliers make the real number a moving target. And yet, total compensation — base salary plus equity plus bonus — is the single biggest factor in most job decisions.
We set out to rank all 35 companies in our Culture Directory by estimated total compensation for mid-to-senior technical roles (typically L5/E5 equivalent). To do this, we cross-referenced data from Levels.fyi, Glassdoor salary reports, Blind discussions, and self-reported compensation data. Where public data was abundant, we marked our confidence as HIGH. Where we were working from fewer data points, we marked it MEDIUM or LOW.
The range you see represents the typical annual total compensation for an experienced engineer or equivalent individual contributor — not entry-level, not VP. Think of it as "what a strong hire with 5–10 years of experience can realistically expect in their first year."
Data Confidence Levels
Not all compensation data is created equal. We've tagged each estimate with a confidence level so you know how much weight to give each number:
- HIGH — Multiple public data points from Levels.fyi, verified Glassdoor reports, and Blind threads. Companies with enough reported offers that the range is well-established.
- MEDIUM — Some data points available, but fewer verified reports. Range is directionally correct but could shift by 10–15%.
- LOW — Limited public data, primarily estimated from peer comparisons, funding stage, and recruiter intelligence. Treat these as educated guesses.
The Full Rankings
Below are all 35 companies sorted by estimated total compensation, from highest to lowest. The table includes Glassdoor rating, WLB score, and our confidence level for each estimate.
| # | Company | TC Range | Glassdoor | WLB | Confidence |
|---|---|---|---|---|---|
| 1 | OpenAI | $350k–$550k | 4.5 | 3.6 | HIGH |
| 2 | DeepMind | $300k–$500k | 4.2 | 4.0 | HIGH |
| 3 | Anthropic | $300k–$490k | 4.4 | 3.7 | HIGH |
| 4 | Databricks | $280k–$450k | 4.1 | 3.9 | MEDIUM |
| 5 | Stripe | $270k–$420k | 4.0 | 3.6 | HIGH |
| 6 | Ramp | $250k–$400k | 4.2 | 3.5 | MEDIUM |
| 7 | Perplexity | $250k–$400k | 4.7 | 3.3 | MEDIUM |
| 8 | CoreWeave | $220k–$380k | 3.6 | 3.2 | MEDIUM |
| 9 | Airbnb | $200k–$380k | 4.1 | 4.0 | HIGH |
| 10 | ElevenLabs | $200k–$350k | 4.2 | 3.6 | MEDIUM |
| 11 | Figma | $200k–$350k | 3.7 | 3.1 | MEDIUM |
| 12 | Notion | $200k–$350k | 4.4 | 4.2 | MEDIUM |
| 13 | Scale AI | $200k–$350k | 3.5 | 2.7 | MEDIUM |
| 14 | Cursor | $200k–$350k | 4.0 | 3.5 | LOW |
| 15 | HubSpot | $180k–$300k | 4.3 | 4.1 | HIGH |
| 16 | Runway | $180k–$320k | 4.5 | 4.0 | MEDIUM |
| 17 | Together AI | $180k–$320k | 4.1 | 3.8 | LOW |
| 18 | Vercel | $180k–$320k | 3.9 | 3.4 | MEDIUM |
| 19 | Pinecone | $180k–$300k | 4.2 | 4.3 | LOW |
| 20 | Modal | $180k–$320k | 4.0 | 3.8 | LOW |
| 21 | Linear | $180k–$320k | 4.6 | 4.4 | LOW |
| 22 | Replit | $170k–$300k | 4.0 | 3.9 | LOW |
| 23 | LangChain | $170k–$300k | 4.6 | 4.0 | LOW |
| 24 | Mistral | $150k–$300k | 4.0 | 3.6 | LOW |
| 25 | Apollo.io | $150k–$280k | 4.0 | 3.6 | MEDIUM |
| 26 | Vast AI | $150k–$280k | 5.0 | 4.5 | LOW |
| 27 | Weaviate | $150k–$280k | 4.3 | 4.2 | LOW |
| 28 | Cohere | $150k–$280k | 2.9 | 3.5 | MEDIUM |
| 29 | Hugging Face | $150k–$280k | 3.8 | 4.1 | LOW |
The spread is enormous. The top-paying company (OpenAI) has a TC ceiling nearly double the bottom tier. But raw compensation is only part of the picture — company stage, equity structure, and location all change what these numbers mean in practice. A $350k package at a pre-IPO company carries very different risk than $300k at a public company like Airbnb.
The Top 5: Where the Money Is
The top five companies represent the apex of AI compensation in 2026. These are the places where senior engineers and researchers can realistically earn $300k+ in total annual compensation. Let's look at each one.
1. OpenAI
OpenAI sits at the top of our compensation rankings, and it's not particularly close. With a TC range of $350k–$550k for senior ICs, the company is paying what it takes to attract the best AI researchers and engineers on the planet. The premium is driven by several factors: intense competition for AI talent, massive funding ($13B+ from Microsoft), and the reality that OpenAI's work is literally defining the industry. The 4.5 Glassdoor rating confirms that employees value the opportunity, even if the 3.6 WLB score suggests they're paying for it with their time.
2. DeepMind
DeepMind offers Google-level compensation with research lab prestige. As part of Alphabet, employees get the stability of a public company's stock (GOOGL) combined with the intellectual freedom of a world-class AI research lab. The $300k–$500k range is well-documented on Levels.fyi, and the 4.0 WLB score is notably higher than most frontier AI labs — Google's cultural infrastructure provides real guardrails against burnout. The trade-off is speed: several reviewers note that Alphabet's bureaucracy can slow things down compared to nimbler competitors.
3. Anthropic
Anthropic is OpenAI's closest competitor in both mission and compensation. The $300k–$490k TC range reflects the company's position as the leading AI safety-focused lab, backed by billions from Amazon and Google. At ~1,500 employees, Anthropic is still small enough to feel like a startup but large enough to pay like a tech giant. The key differentiator from OpenAI: Anthropic's explicit focus on AI safety attracts a specific type of engineer who cares about responsible development. The 4.4 Glassdoor rating and 3.7 WLB score both sit slightly above OpenAI's WLB, suggesting a marginally more sustainable pace.
4. Databricks
Databricks has the most open roles of any company in our database (791 jobs), and the compensation matches the hiring appetite. At a $62B valuation, the company pays like a late-stage pre-IPO giant — generous base salaries plus equity that could see significant upside if the company goes public. The MEDIUM confidence tag reflects the fact that Databricks compensation is well-known at senior levels but less documented for mid-career roles. The 3.9 WLB score suggests a company that works hard but hasn't crossed into burnout territory.
5. Stripe
Stripe is perhaps the most well-documented company on our list when it comes to compensation. With thousands of data points on Levels.fyi and a reputation for methodical, writing-driven culture, the $270k–$420k range is backed by HIGH confidence. Stripe also stands out for its engineering culture: the company is famous for its emphasis on clear writing, code quality, and infrastructure thinking. At ~8,000 employees, it's the largest company in the top five, proving that scale and strong compensation can coexist. The 3.6 WLB score is typical of high-growth fintech — the work is demanding but the craft is respected.
How We Estimated These Numbers
Compensation data in tech is notoriously opaque. Here's our methodology, warts and all:
- Levels.fyi was our primary source. For companies with 20+ verified data points (OpenAI, Stripe, Airbnb, HubSpot, DeepMind), we used the reported ranges directly. These are marked HIGH confidence.
- Glassdoor salary reports provided a second data point, though Glassdoor tends to undercount equity, making their total comp figures lower than reality for equity-heavy companies.
- Blind discussions gave us directional data for private companies where official comp data is scarce. We treated these as qualitative signals, not precise numbers.
- Peer comparison was used for smaller companies (Cursor, Vast AI, Modal) where public data is minimal. If a Series B AI company is competing for the same talent as Anthropic, its offers need to be in the same ballpark — we estimated accordingly.
- Location adjustment: All ranges assume US-based compensation, typically San Francisco or New York. Companies like Mistral (Paris-based) may have different local market rates, which is why their range carries a LOW confidence tag.
We will update these estimates quarterly as new data becomes available. If you have verified compensation data from any of these companies, we'd love to hear from you.
Equity: The Wildcard
The biggest variable in these numbers is equity. For public or late-stage companies like Airbnb, Stripe, and HubSpot, equity is liquid or near-liquid — you can reasonably count it as part of your annual compensation. For early-stage companies, equity is a bet.
Consider the extremes. Cursor has ~50 employees and a TC range of $200k–$350k. If you joined early and the company hits a $10B+ valuation, your equity could be worth multiples of your base salary. But if the company stumbles, that equity is worth nothing. Compare that to HubSpot, where your $180k–$300k includes RSUs in a publicly traded company (NYSE: HUBS) — less upside, but it's real money you can sell today.
The companies in the middle — Databricks, Ramp, Perplexity — are the most interesting from an equity perspective. They're valued high enough that equity feels substantial, but they're still private, meaning you're trusting the valuation until a liquidity event. If you're evaluating offers from these companies, the question isn't "how much is the equity worth?" but rather "how confident am I in the next 3–5 years of this company?"
We track which companies in our database emphasize equity as a core value — meaning they give meaningful ownership stakes, not just token option grants. Browse those companies to find places where your upside is real.
The Trade-Offs: Pay vs. Everything Else
High compensation doesn't exist in a vacuum. The data reveals clear trade-offs between pay and other dimensions of workplace quality:
- Pay vs. WLB: The top 5 highest-paying companies have an average WLB score of 3.76. The bottom 5 average 4.06. Higher pay often correlates with more demanding cultures.
- Pay vs. Company Size: Small companies (<200 employees) tend to cluster at the bottom of the TC rankings, not because they can't afford to pay, but because they have less compensation data available. The LOW confidence tags are disproportionately assigned to small companies.
- Pay vs. Mission: Frontier AI labs (OpenAI, Anthropic, DeepMind) dominate the top of the rankings because they're competing for a narrow pool of AI researchers. Developer tools companies (Vercel, Linear, Replit) pay well but can't match the pure-play AI labs.
The best advice we can give: don't optimize for compensation alone. A $350k package at a company with a 2.7 WLB score might leave you burned out in 18 months. A $200k package at a company with a 4.4 WLB score and strong equity upside might be the better long-term play. Use our comparison tool to evaluate companies across all dimensions, not just pay.
Browse all 5,074 jobs from 35 companies
Find your next role — and filter by the culture values that matter to you.
Browse All Jobs → Equity-First Companies →