Skip to main content
Live
Main content

Fireworks AI co-founder Chenyu Zhao says most companies can skip frontier models

Speaking at Slush, Zhao put open models three to six months behind the frontier as Fireworks works through two security incidents this fall.

Jaeden Schafer
Editor in Chief · · 5 min read
Fireworks AI co-founder Chenyu Zhao says most companies can skip frontier models

Fireworks AI co-founder Chenyu Zhao said on Thursday that most companies can run the bulk of their AI workloads on open-weight models instead of paying for frontier models from OpenAI and Anthropic. Speaking on stage at the Slush conference in Helsinki, Zhao said open models are at most six months behind the frontier, and probably closer to three.

Fireworks AI runs AI models for other businesses and specializes in open-weight models, with customers that have included Uber, Shopify and Notion. Zhao described Fireworks as a specialized intelligence platform, set against the general-purpose systems the big labs are building.

“OpenAI and Anthropic are building general intelligence.”
— Chenyu Zhao, Fireworks AI co-founder

Zhao laid out when he thinks the switch makes sense. Before a company finds product-market fit, using the best frontier model is low risk, he said, because those models are a household brand. Once a product reaches 10,000 to 100,000 users and is spending $10,000 to $100,000 a month on tokens, it is time to consider open source.

Key facts

  • 01Fireworks AI co-founder Chenyu Zhao said open-weight models are at most six months behind frontier models, probably closer to three.
  • 02Zhao said companies spending $10,000 to $100,000 a month on tokens should consider open-source models.
  • 03Investors valued Fireworks at $17.5 billion this summer; the company is reportedly in early talks for another round.
  • 04Fireworks disclosed two security incidents this fall; Notion has disconnected from the platform.
  • 05Uber says about 150,000 drivers' license details were stolen through a third-party AI vendor.

He argued that relying on a frontier model alone does not set a product apart, saying that just using Opus or Sonnet is not a differentiator. Zhao said he expects a two-model world, with frontier and open models running in parallel, and noted that companies do not need a frontier model for basic arithmetic.

Investors valued Fireworks at $17.5 billion this summer, and the company is reportedly in early talks for another round. Zhao declined to comment on fundraising. Rivals Together AI and Baseten have announced large rounds in recent months, and on Wednesday Palo Alto Networks agreed to buy Applied Compute, founded a year ago by former OpenAI researchers, for $4.5 billion in stock.

Zhao spent a decade at Google, including managing a 50-person engineering team on Google Cloud Vertex AI, before co-founding Fireworks with Lin Qiao and Dima Dzhulgakov. His remarks came the same day Anthropic went public, and Zhao said he would probably call his broker that night to buy a few shares, calling it a long-term bet.

Fireworks AI has disclosed two security incidents this fall. In the first, hackers reached API keys and tokens stored as secrets for a limited number of accounts. In the second, attackers breached Fireworks' cloud provider, got into a development system and took customer data, including billing records and some datasets.

Notion says AI chat transcripts from fewer than 1% of its workspaces were exposed, and it has disconnected from Fireworks. Uber says hackers stole the names, addresses, birth dates and driver's license numbers of about 150,000 drivers through a third-party vendor that has been identified as Fireworks. Zhao said he could not publicly name the specific customers affected.

“Every company is probably seeing an unprecedented frequency of attacks.”
— Chenyu Zhao, Fireworks AI co-founder
Related · from this week
HPE argues enterprises should own AI capacity as production workloads scale
Jaeden Schafer · 4 min read →

Zhao said Fireworks promises zero data retention for inference traffic and that there is no indication that promise has been compromised. He said the affected Notion workload used batch prediction, which stores uploaded datasets at rest, and that Fireworks is moving those services to an online flow. He added that much of the motivation behind the attack was LLM jacking.

Nvidia has backed Fireworks since 2024. Asked about Jensen Huang's departure from Nvidia, Zhao called it a blow to the industry and said Nvidia has always been a strong supporter of open source, adding that he does not see that changing.

ShareXLinkedInEmail
AI Box

Every AI model. One chat.

The latest models from ChatGPT, Claude, Gemini, Sora, ElevenLabs — 80+ models in a single chat. Compare answers side by side. Pick the best one every time.

  • ChatGPT, Claude, Gemini, Grok, DeepSeek — in one chat
  • Generate images & video with Sora, Veo, Ideogram
  • Compare any two models side by side
  • From $8.99/mo · 80+ models, all included
Try AI Boxaibox.ai
Trusted by 3,000+ teams
Got a tip?

Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.

Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.

AI Box Daily briefingFree · Daily · No fluff

Stay ahead of everyone in AI.

The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.

Loved by 10,000+ AI professionals
Free forever. Unsubscribe with one click.

The briefing read inside teams at

Keep reading

More from Business

HPE argues enterprises should own AI capacity as production workloads scale
Business

HPE argues enterprises should own AI capacity as production workloads scale

Deloitte data shows the share of firms with 40% of AI projects in production is set to double within six months, changing the cost math.

Jaeden Schafer4 min read
Nvidia logo
Business

China weighs letting ByteDance and Alibaba buy new Nvidia chips

A policy reversal would reopen one of Nvidia's largest export markets after months of Beijing steering domestic buyers toward local silicon.

Jaeden Schafer4 min read
Blue Cross Blue Shield says AI hospital coding added $942M in costs
Business

Blue Cross Blue Shield says AI hospital coding added $942M in costs

A BCBSA analysis blames AI-assisted claims tools for a two-year spending surge, with no matching change in care delivered.

Jaeden Schafer4 min read