Is Anyscale profitable in 2026?

No audited figures are public. Anyscale said in September 2024 it was 'on track to quadruple revenue' that year (vendor claim), but third-party revenue trackers range from $24M to $380M for 2023-2024, so profitability cannot be confirmed.

Anyscale is privately held and venture-backed, founded in 2019 by Ray creators Robert Nishihara, Philipp Moritz, Ion Stoica and UC Berkeley's Michael I. Jordan. It was last valued at $1B after its December 2021 Series C; no new priced round is confirmed as of mid-2026.

Could Databricks replace Anyscale?

Databricks, valued at $134B after a December 2025 Series L round, is the most frequently cited adjacent competitor and is far larger than Anyscale's last disclosed $1B (2021/2022) valuation, pressuring Anyscale's entry-level enterprise AI-infrastructure market.

What is Anyscale's biggest competitor?

Databricks is the most cited adjacent giant, valued at $134B in December 2025, alongside hyperscaler-native services like AWS SageMaker, Azure ML and Google Vertex that compete for the same Ray-adjacent enterprise AI-infrastructure budgets.

Has Anyscale had layoffs?

No layoffs are documented in available records as of mid-2026. Anyscale instead sunset its Endpoints inference product in late 2024 and has continued hiring, including a Field CTO appointment in March 2026 to scale enterprise Ray adoption.

Could AI agents replace Anyscale's platform?

Unlikely near-term; Anyscale is positioning itself as AI-agent-native infrastructure, launching 'Agent Skills' on April 22, 2026, letting AI coding agents like Claude and Copilot directly debug and deploy Ray workloads rather than being displaced by them.

AI companies and labs · United States

Anyscale

anyscale.com ↗

Anyscale sells managed infrastructure and support for Ray, the open source distributed compute framework for AI.

Losing its protection

Assessed 20 July 2026 · Evidence as of 20 July 2026

The structural protection is eroding.

On the recordSince 20 July 2026 · Checkable from H2 2026 · Every call on the record Checkpoint: Azure managed Ray general availability, currently in public preview

The call

Anyscale controls the commercial home of a widely used open source framework, but losing exclusive governance and facing Databricks head on with a four year stale valuation makes this a company proving itself, not one that has already proven it.

Anyscale built Ray, an open-source framework downloaded over 237 million times and used by companies like Uber and BMW to process large amounts of data quickly. It was once hard to replace because Anyscale owned it exclusively, but in October 2025 Ray moved to neutral governance, meaning anyone can develop it freely. Now Databricks (a much larger company worth $134 billion) and cloud services from Amazon, Microsoft, and Google are offering similar tools directly, making Ray less special. The one thing that could change everything is if Microsoft or Databricks convince Anyscale's biggest customers to switch to their managed versions instead, which could happen as soon as the second half of 2026.

Why

AI pressure: None apparent. AI workloads are Anyscale's customer base, not its predator; growth in AI training/inference demand feeds Ray usage rather than eating it. Could AI do the core job? Unlikely.

Bigger players: Databricks and hyperscaler-native services (AWS SageMaker, Azure ML, Google Vertex AI). Databricks ($134B) and cloud providers can ship managed Ray natively now that governance moved to PyTorch Foundation in Oct 2025, undercutting Anyscale's commercial layer. Could a giant copy it? Yes, easily.

What protects it. Existing enterprise lock-in at Uber, BMW, Adyen and Coinbase running production Ray workloads that are costly to migrate off Anyscale's managed layer.

What would change this call

Azure ships GA managed Ray service in H2 2026. Would show Microsoft is converting Anyscale's own distribution partner into a direct competitor.
BMW, Adyen, or Coinbase publicly move workloads to Databricks or a hyperscaler-native Ray offering. Would show Anyscale's marquee accounts are churning to rivals.
Anyscale announces a new priced funding round with disclosed valuation. Would confirm whether revenue growth claims are real or the $1B mark is stale.

Key numbers

Ray hit 237M cumulative downloads (Oct 2025)
Valuation stuck at $1B since Series C (Aug 2022)
Revenue estimates range $24M to $380M across trackers (2026)

Named risks

No new priced funding round confirmed since 2022, valuation stuck near $1B. Severity: high
Loss of exclusive Ray governance after PyTorch Foundation move, Oct 2025. Severity: high
Prior product retreat precedent with Endpoints sunset in late 2024. Severity: medium
Wildly inconsistent third-party revenue and headcount estimates undermine traction claims. Severity: medium
ShadowRay vulnerability (CVE-2023-48022) still cited as ongoing enterprise security concern. Severity: low

Rivals, as read in July 2026

Rival	Direction	Evidence
Databricks	losing to	Valued $134B Dec 2025 vs Anyscale's stale $1B mark from 2021/2022
AWS SageMaker, Azure ML, Google Vertex	contesting	Hyperscaler-native managed AI services unblocked to build competing Ray-like offerings
Modal, Together AI, Fireworks AI	gaining on	Serverless GPU/inference specialists absorbed demand after Anyscale sunset Endpoints in late 2024
Microsoft (Azure)	contesting	Partner since Nov 2025 private preview but could natively replicate managed Ray post GA

Questions people ask

Is Anyscale profitable in 2026?: No audited figures are public. Anyscale said in September 2024 it was 'on track to quadruple revenue' that year (vendor claim), but third-party revenue trackers range from $24M to $380M for 2023-2024, so profitability cannot be confirmed.
Who owns Anyscale?: Anyscale is privately held and venture-backed, founded in 2019 by Ray creators Robert Nishihara, Philipp Moritz, Ion Stoica and UC Berkeley's Michael I. Jordan. It was last valued at $1B after its December 2021 Series C; no new priced round is confirmed as of mid-2026.
Could Databricks replace Anyscale?: Databricks, valued at $134B after a December 2025 Series L round, is the most frequently cited adjacent competitor and is far larger than Anyscale's last disclosed $1B (2021/2022) valuation, pressuring Anyscale's entry-level enterprise AI-infrastructure market.
What is Anyscale's biggest competitor?: Databricks is the most cited adjacent giant, valued at $134B in December 2025, alongside hyperscaler-native services like AWS SageMaker, Azure ML and Google Vertex that compete for the same Ray-adjacent enterprise AI-infrastructure budgets.
Has Anyscale had layoffs?: No layoffs are documented in available records as of mid-2026. Anyscale instead sunset its Endpoints inference product in late 2024 and has continued hiring, including a Field CTO appointment in March 2026 to scale enterprise Ray adoption.
Could AI agents replace Anyscale's platform?: Unlikely near-term; Anyscale is positioning itself as AI-agent-native infrastructure, launching 'Agent Skills' on April 22, 2026, letting AI coding agents like Claude and Copilot directly debug and deploy Ray workloads rather than being displaced by them.

Timeline

2026-06-02Anyscale on Azure moved to public preview, drawing down Azure MACC commitments
2026-04-22Anyscale launched 'Agent Skills' for AI-agent-native Ray debugging and deployment
2026-03-19Christian Stano hired as Field CTO to scale enterprise Ray adoption
2025-12Databricks valued at $134B after Series L round (competitor context)
2025-11Microsoft Azure managed-Ray partnership first announced, private preview
2025-10-22Ray joined Linux Foundation's PyTorch Foundation, ending exclusive Anyscale governance
2024-07Keerti Melkote appointed Anyscale CEO, succeeding co-founder leadership
2024Anyscale Endpoints hosted LLM inference product sunset
2021-12Series C, $100M raised at $1B valuation, co-led by a16z and Addition
2020-10Series B, $40M raised, led by NEA

Revision history

20 July 2026

First assessed. Losing its protection. The text above is fixed as of this date and is not edited afterwards.

In the same band

Safe Superintelligence Inc.Mistral AI Cohere