AI/ML Salary Bands - Know Your Worth
Reading time: ~18 min | Interview relevance: High | Roles: All
The Real Interview Moment
You just passed the final round. The recruiter calls: *"We'd like to extend an offer. The base is 185K sounds great. But you don't know that the L5 band for this role at this company goes up to 80K/year, and the signing bonus is negotiable up to 100K+ on the table.
This page gives you the data to never make that mistake. Salary transparency is your most powerful negotiation tool.
What You Will Master
- Total compensation structure (base + bonus + equity + signing)
- Salary bands by role, level, company tier, and geography
- How to research your market rate using levels.fyi, Glassdoor, and Blind
- Equity types and how to value them (RSUs, ISOs, startup equity)
- When and how to use this data in negotiations
- Common traps in AI/ML compensation
Part 1 - How AI/ML Compensation Works
Total Compensation (TC) Breakdown
| Component | What It Is | Typical Range (L5) | Key Details |
|---|---|---|---|
| Base Salary | Fixed cash, paid biweekly | $170\text{--}230K | Least negotiable at big tech, most negotiable at startups |
| Equity (RSUs) | Stock units vesting over 4 years | $80–200K/year | Valued at grant price; can go up or down |
| Annual Bonus | Performance-based cash | 15–25% of base | Target %, actual depends on performance + company |
| Signing Bonus | One-time cash at start | $20\text{--}80K | Most negotiable component; used to bridge equity gaps |
Compensation conversations happen after the interview. But your interview performance directly affects your leverage. A "Strong Hire" rating across all rounds gives the recruiter room to push for top-of-band offers. A "Lean Hire" gets you the minimum of the band. Your interview prep IS your salary negotiation prep.
Part 2 \text{---} Salary Bands by Role and Level
Big Tech (FAANG-Tier) \text{---} US, 2026
All figures are total compensation (TC) in USD. Ranges represent the full band from new-in-level to top-of-band.
| Level | MLE | AI Engineer | MLOps | Data Scientist | Research Eng. | Data Engineer |
|---|---|---|---|---|---|---|
| Junior (L3) | $150\text{--}250K | $130\text{--}220K | $130\text{--}210K | $120\text{--}200K | $150\text{--}250K | $120\text{--}190K |
| Mid (L4) | $200\text{--}350K | $180\text{--}320K | $180\text{--}300K | $170\text{--}280K | $220\text{--}380K | $170\text{--}270K |
| Senior (L5) | $300\text{--}450K | $250\text{--}400K | $250\text{--}380K | $220\text{--}350K | $280\text{--}420K | $220\text{--}340K |
| Staff (L6) | $400\text{--}600K | $350\text{--}550K | $350\text{--}500K | $320\text{--}450K | $380\text{--}550K | $300\text{--}450K |
| Principal (L7) | $500\text{--}800K | $450\text{--}700K | $450\text{--}650K | $400\text{--}550K | $450\text{--}650K | $400\text{--}550K |
AI Startups (Well-Funded, Series B+) \text{---} US, 2026
| Level | MLE | AI Engineer | MLOps |
|---|---|---|---|
| Mid (L4-equiv) | $180\text{--}280K + equity | $160\text{--}260K + equity | $160–240K + equity |
| Senior (L5-equiv) | $220\text{--}350K + equity | $200–320K + equity | $200\text{--}300K + equity |
| Staff (L6-equiv) | $280\text{--}400K + equity | $260\text{--}380K + equity | $260–350K + equity |
Startup equity is highly variable. See the equity section below for how to evaluate it.
Frontier AI Labs - US, 2026
These labs (OpenAI, Anthropic, DeepMind) pay premium compensation:
| Level | Research Engineer | AI Engineer / MLE |
|---|---|---|
| Mid | $250\text{--}400K | $220–380K |
| Senior | $350\text{--}550K | $300–500K |
| Staff | $500\text{--}750K | $450–650K |
Geographic Adjustments
| Location | Adjustment vs. SF/NYC |
|---|---|
| San Francisco / NYC | Baseline (100%) |
| Seattle / LA | 95–100% |
| Austin / Denver / Boston | 85–95% |
| Other US cities | 75–90% |
| London | 70–85% |
| Berlin / Amsterdam | 60–75% |
| Bangalore / Hyderabad | 30–45% |
| Remote (US-based company) | 80–100% (depends on company policy) |
Don't compare base salaries across companies - compare total compensation. A startup offering 180K base + 30K bonus. Always calculate the full TC before comparing offers.
Part 3 - How to Research Your Market Rate
Data Sources (Ranked by Reliability)
| Source | Reliability | Best For | Limitations |
|---|---|---|---|
| levels.fyi | ⭐⭐⭐⭐⭐ | Big tech TC by company/level/role | Less data for startups |
| Blind | ⭐⭐⭐⭐ | Anonymous TC sharing, negotiation advice | Skews high (survivorship bias) |
| Glassdoor | ⭐⭐⭐ | Broad coverage, base salary data | TC data often incomplete |
| comp.fyi | ⭐⭐⭐⭐ | Startup compensation + equity data | Newer, less data |
| H1B Salary Database | ⭐⭐⭐⭐ | Verified base salaries for visa holders | Base only, no equity/bonus |
| Your network | ⭐⭐⭐⭐⭐ | Specific, current, contextual | Requires strong network |
Research Workflow
- Identify your target: Role + Level + Company (or company tier)
- Check levels.fyi: Search for your exact role/level/company. Look at the range, not just the median.
- Cross-reference on Blind: Search for recent offer posts at that company. Note the negotiation stories.
- Adjust for geography: Apply the geographic multiplier if you're not in SF/NYC.
- Calculate your BATNA: Your Best Alternative to Negotiated Agreement. What's your next-best offer?
"To know my market rate, I'd check levels.fyi for TC data at my target company and level, cross-reference with Blind for recent offers, and adjust for location. I'd also calculate the full TC breakdown - base, equity, bonus, signing - because comparing base salary alone is misleading. Most importantly, I'd have multiple offers to establish a real BATNA before negotiating."
Part 4 - Understanding Equity
RSUs vs. Options vs. Startup Equity
| Type | What You Get | Risk Level | Common At |
|---|---|---|---|
| RSUs | Actual shares that vest over time | Low (worth current stock price) | Big tech, public companies |
| ISOs | Right to buy shares at a strike price | Medium (worth $0 if stock drops below strike) | Late-stage startups |
| NSOs | Same as ISOs but taxed as income | Medium | Mid-stage startups |
| Startup equity | % ownership in a private company | High (could be worth $0 or millions) | Early-stage startups |
How to Value Startup Equity
Startups will tell you "your equity is worth $X based on our latest valuation." This is misleading. You can't sell private shares. Apply a significant discount: 50-75% for Series A, 25-50% for Series B+. And always ask: What's the preference stack? In a down-round acquisition, preferred shareholders (investors) get paid first - common shareholders (employees) may get nothing.
Part 5 - Level Calibration Across Companies
The same title means different things at different companies:
| Company | Junior | Mid | Senior | Staff | Principal |
|---|---|---|---|---|---|
| L3 | L4 | L5 | L6 | L7 | |
| Meta | E3 | E4 | E5 | E6 | E7 |
| Amazon | SDE1 | SDE2 | SDE3 / Sr SDE | Principal | Sr Principal |
| Apple | ICT2 | ICT3 | ICT4 | ICT5 | ICT6 |
| Microsoft | 59-60 | 61-62 | 63-64 | 65-66 | 67-68 |
| Netflix | - | - | Senior | Staff | - |
| OpenAI | L2 | L3 | L4 | L5 | L6 |
| Startups | Junior | Engineer | Senior | Staff / Lead | Principal / CTO |
Netflix famously only hires at senior+ level and pays top-of-market base with no equity. Amazon front-loads equity vesting: 5% year 1, 15% year 2, 40% each in years 3-4 - so your TC jumps significantly in year 3. Google has the most uniform vesting (25% per year). Always ask about the vesting schedule.
Practice Problems
Problem 1: Offer Evaluation
You have two offers:
- Company A (Big Tech): 150K/year RSUs (4-year vest), 15% bonus target, $40K signing bonus
- Company B (Series B Startup): 500M), $20K signing bonus
Which is worth more?
Hint 1 - Direction
Calculate total year-1 compensation for Company A. For Company B, discount the startup equity significantly.
Full Answer + Rubric
Company A Year-1 TC: 37.5K RSUs (25% of 28.5K bonus (15% of 40K signing = $296K
Company B Year-1 TC: 500M = 125K\text{--}250K over 4 years → 20K signing = $261K–292K
But that's not the full picture:
- Company A RSUs are in publicly traded stock - liquid and relatively low-risk.
- Company B equity is illiquid, subject to dilution, and could be worth $0 if the company fails.
- Risk-adjusted, Company A is worth more.
- However, if Company B succeeds and reaches $5B valuation, your equity would be 10x. It's a gamble.
Scoring:
- Strong Hire: Calculates both correctly, risk-adjusts the startup equity, considers upside scenario
- Lean Hire: Gets the math right but doesn't risk-adjust
- No Hire: Compares base salaries only
Problem 2: Negotiation Strategy
You received an offer for 400K for your level. How do you negotiate?
Full Answer + Rubric
Strong approach:
- Thank them and express genuine interest. Don't negotiate aggressively on the first call.
- Ask for the full breakdown: base, equity, bonus, signing. Understand which components are flexible.
- Present data: "I've researched the market rate for this role and level, and competitive offers are in the $350-380K range. Is there flexibility in the equity or signing bonus?"
- Use competing offers (if you have them): "I have another offer at $360K TC. I'd prefer to join your team - can we close the gap?"
- Negotiate components separately: Equity is usually most flexible. Signing bonus second. Base is least flexible at big tech.
- Know your bottom line: What's the minimum you'd accept? Walk away if the offer doesn't meet it.
Key phrases: "Is there flexibility?", "What would it take to reach X?", "I'm excited about this role - I want to make this work."
Scoring:
- Strong Hire: Data-driven, negotiates components separately, uses competing offers, knows walk-away number
- Lean Hire: Reasonable approach but doesn't have specific data or competing offers
- No Hire: Accepts first offer or makes aggressive demands without data
Interview Cheat Sheet
| Situation | What to Do | What NOT to Do |
|---|---|---|
| "What are your salary expectations?" | "I'm targeting market rate for this role and level. Based on my research, that's $X-Y TC." | Give a single number (anchors you low) |
| "What's your current salary?" | "I'd prefer to discuss the value I bring to this role." (It's illegal to ask in many US states) | Disclose if you're currently underpaid |
| "This is our best and final offer" | "I appreciate that. Can we discuss the equity or signing bonus specifically?" | Accept immediately or get adversarial |
| You have multiple offers | "I have competing offers in the $X range. I'd prefer your team - can we close the gap?" | Lie about offers (recruiters talk to each other) |
| You have no other offers | Still negotiate. "Based on market data, I'd like to discuss..." | Mention you have no alternatives |
Spaced Repetition Checkpoints
- Day 0: Read this page. Check levels.fyi for your target role/company/level. Note the range.
- Day 3: Calculate the TC for a sample offer: base + equity + bonus + signing.
- Day 7: Research 3 companies' compensation structures. Note vesting schedules and bonus targets.
- Day 14: Practice the negotiation script with a friend. Record yourself and critique.
- Day 21: Review your BATNA. Do you have multiple offers (or potential offers) to strengthen your position?
What's Next
- For negotiation tactics → Negotiation & Offers
- To understand levels better → Career Ladders
- Back to role selection → Overview