The AI Race: Billions Spent, But Who’s Really Winning in 2025?

Aakash Gupta
2 min readApr 2, 2025

--

The leading AI foundational model landscape continues to evolve at a dizzying pace. First o3-mini took the spotlight, then Claude 3.7 Sonnet, and now Gemini-2.5 pro has entered the arena. Here’s my breakdown of where the major players stand today and what it means for the industry.

The Contenders

Here’s who’s competing in the foundation model space right now:

  • OpenAI — o3-mini
  • Anthropic — Claude 3.7 Sonnet Thinking
  • xAI — Grok 3
  • DeepSeek — R1, MoE model
  • Google — Gemini 2.5 Flash
  • Meta — LLaMA 3.4 Turbo

Benchmarks: A New Leader Emerges

Just last week, Claude and OpenAI were neck and neck for the leadership position:

  • Claude → 76.10
  • o3-mini → 75.88

Now, Google has dramatically reshuffled the deck:

  • Gemini-2.5 pro → 82.35

This represents a significant leap forward, with Google taking the lead across virtually all benchmark categories.

Fundraising: Follow the Money

Here’s where things get particularly interesting. Let’s examine fundraising and estimated spending:

  • OpenAI: $57.9B raised
  • Google: $50B spent
  • Meta: $37B spent
  • Anthropic: $14.3B
  • xAI: $6B raised
  • DeepSeek: $0.6B

The bottom line? OpenAI is effectively the Uber of this space, with Anthropic playing the role of Lyft. However, the sheer financial power of Google and Meta makes them formidable competitors that shouldn’t be underestimated.

Revenue in 2024: The Business Reality

This is where OpenAI truly separates itself from the pack:

  • OpenAI: $3.5B
  • Google: $2.5B (estimated)
  • Anthropic: $1B
  • xAI: $0.1B
  • DeepSeek: $0
  • Meta: $0 (no direct AI revenue yet)

The Uber comparison becomes even more apparent when looking at these revenue figures. OpenAI has established a clear lead in monetizing its AI technologies.

Headcount: Size Matters

This metric reveals Google’s significant advantage:

  • Google: ~6,000 (DeepMind/Research)
  • OpenAI: ~3,350
  • Meta: ~2,500
  • Anthropic: ~1,035
  • xAI: ~1,000
  • DeepSeek: ~125

More people typically means more experiments and greater output. Google’s substantial research team gives them considerable firepower for innovation.

Search Trends: The Popularity Contest

Sometimes internet trends tell the most important part of the story:

  • ChatGPT still dominates search interest
  • Claude is steadily gaining momentum
  • Gemini, Llama, and DeepSeek remain far behind in terms of public attention

This matters because in the AI space, mindshare is increasingly translating into market share.

Career Opportunities

Interested in working at one of these industry leaders?

The Big Picture

It’s not enough to simply “know AI” in today’s landscape. To truly understand what’s happening, you need to:

  1. Comprehend the broader ecosystem
  2. Grasp the business dynamics at play
  3. Understand the distribution strategies being employed

Consider this your cheat sheet to navigating the rapidly evolving world of foundation AI models in 2025.

Read my newsletter here.

What are your thoughts on the current AI landscape? Share your perspective in the comments below.

--

--

Aakash Gupta
Aakash Gupta

Written by Aakash Gupta

Helping PMs, product leaders, and product aspirants succeed

No responses yet