Cross-Provider Analysis: OpenAI & Anthropic — Valuations, Secondary Markets, and IPO Timelines (April 6, 2026)

Executive Summary

• OpenAI completed the largest private tech funding round in history — $122 billion at an $852 billion post-money valuation ^[50] — yet secondary markets immediately revealed a ~10% discount, with bids implying only ~$765 billion and ~$600 million in employee shares struggling to find buyers ^[3]. No formal "down round" has occurred, but secondary pricing signals meaningful investor skepticism at the primary valuation.

• Anthropic is the secondary market darling: over $2 billion in buy orders — including $1.6 billion on Hiive alone — valued Anthropic at ~$600 billion ^[3], a 50%+ premium to its February 2026 Series G primary valuation of $380 billion ^[2]. Goldman Sachs charges full 15–20% carry for Anthropic access while waiving fees to attract OpenAI buyers — a stark signal of relative institutional demand ^[2].

• Revenue trajectories are converging dangerously for OpenAI: Anthropic's ARR surged from ~$14 billion in February to ~$19 billion by early March 2026 ^[2], while OpenAI runs at ~$24 billion annualized ^[2]. Anthropic's enterprise API market share rose from 12% to 32% as OpenAI's fell from ~50% to 25% ^[29], and some analyses project Anthropic could surpass OpenAI in annualized revenue by mid-2026 ^[54].

• Both companies face existential cash burn: OpenAI projects ~$14 billion in net losses for 2026 alone ^[2], with cumulative losses potentially reaching $44 billion through 2028 and $200 billion in infrastructure spend committed through 2030 ^[2]. Anthropic's $80 billion cloud-spend commitment through 2029 ^[89] and revised-down gross margin outlook ^[62] signal comparable structural pressure.

• IPO timelines are contested and uncertain: Anthropic is the more likely near-term candidate, with Q4 2026 discussions underway and bankers projecting a $60 billion+ raise ^[3]. OpenAI faces an internal CEO-CFO rift — Sam Altman targets Q4 2026 while CFO Sarah Friar advocates for 2027 ^[3] — and prediction markets assign only 61% probability that either company IPOs by end of 2027 ^[43].

Cross-Provider Consensus

1. OpenAI's $122 Billion Funding Round at $852 Billion Valuation

Providers: Gemini-Lite, Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

All four providers independently confirmed that OpenAI closed a ~$122 billion funding round at a ~$852 billion post-money valuation ^[50], described as the largest private tech funding round in history. Key investors included SoftBank, Amazon, and Nvidia ^[47]. Perplexity added granular detail: Amazon committed $50 billion ($15B upfront, $35B conditional on milestones), Nvidia provided $30 billion, and SoftBank pledged $30 billion in three tranches ^[47].

2. Anthropic's Series G: $30 Billion at $380 Billion Valuation

Providers: Gemini-Lite, Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

All four providers confirmed Anthropic's February 2026 Series G raised $30 billion at a $380 billion post-money valuation ^[4], led by GIC and Coatue, with co-investors including D.E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ, and MGX ^[17]. This represented a more than doubling from the September 2025 valuation of $183 billion ^[58].

3. OpenAI Secondary Market Discount (~$765 Billion Implied Valuation)

Providers: Gemini-Lite, Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

All four providers confirmed that approximately $600 million in OpenAI shares were offered on secondary markets and struggled to find buyers, with bids implying a valuation of approximately $765 billion — roughly a 10% discount to the primary round ^[4]. Grok and OpenAI added the specific detail that Goldman Sachs reportedly waived its usual fees to attract buyers ^[2].

4. Anthropic Secondary Market Premium (~$600 Billion Implied Valuation)

Providers: Gemini-Lite, Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

All four providers confirmed that Anthropic's secondary market demand was exceptional, with bids implying ~$600 billion — a 50%+ premium to the $380 billion primary ^[3]. Hiive alone logged $1.6 billion in demand ^[3]. Goldman Sachs continued charging standard 15–20% carry for Anthropic access ^[2].

5. OpenAI's ~$14 Billion Projected Net Loss in 2026

Providers: Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

Three providers confirmed OpenAI's internal forecasts project approximately $14 billion in net losses for 2026 ^[3], driven by massive infrastructure and compute commitments. OpenAI added that cumulative losses through 2028 are projected at ~$44 billion, with profitability not expected until 2029 ^[2].

6. Anthropic ARR Surge: ~$14 Billion (February) to ~$19 Billion (March 2026)

Providers: Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

Three providers confirmed Anthropic's rapid ARR growth, from ~$14 billion in February 2026 ^[3] to ~$19 billion by early March 2026 ^[2], driven substantially by Claude Code, which achieved a $2.5 billion+ run-rate ^[2].

7. OpenAI CEO-CFO Divergence on IPO Timing

Providers: Gemini-Lite, Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

All four providers confirmed the internal tension: Sam Altman targets a Q4 2026 IPO ^[4], while CFO Sarah Friar has raised concerns about organizational readiness and advocates for a 2027 timeline ^[4].

8. Anthropic Targeting Q4 2026 IPO

Providers: Gemini-Lite, Grok-Premium, OpenAI, Perplexity | Confidence: HIGH

All four providers confirmed Anthropic is in discussions for an IPO as early as Q4 2026 ^[4], with bankers projecting a raise exceeding $60 billion ^[3] and a public valuation in the $400–500+ billion range. Legal counsel Wilson Sonsini has been engaged, and Goldman Sachs, JPMorgan, and Morgan Stanley are in discussions ^[2].

9. No Formal Down Round for Either Company

Providers: Gemini-Lite, Grok-Premium | Confidence: HIGH

Both providers that addressed this point explicitly confirmed there is no evidence of a formal down round for either OpenAI or Anthropic ^[8]. The secondary market discount for OpenAI reflects liquidity and sentiment dynamics, not a restructured primary valuation.

10. OpenAI's $200 Billion Infrastructure Spend Commitment Through 2030

Providers: OpenAI, Perplexity | Confidence: HIGH

Both providers confirmed OpenAI has committed to spend $200 billion on AI supercomputers and model training through 2030 ^[4], though Perplexity noted earlier projections were as high as $1.4 trillion before being revised down to $600 billion ^[56].

Unique Insights by Provider

Gemini-Lite

• Bankers estimate Anthropic's IPO valuation at $400–500 billion: While all providers noted the IPO discussions, Gemini-Lite specifically surfaced the banker consensus range for Anthropic's public market valuation, which is notably below the $600 billion secondary market implied price — suggesting bankers are pricing in a discount to secondary enthusiasm. This gap matters for investors evaluating whether secondary bids are reliable price discovery or speculative froth.

• Market preference rooted in enterprise revenue mix: Gemini-Lite explicitly framed the secondary market divergence as a function of Anthropic's enterprise-heavy revenue mix versus OpenAI's consumer-heavy model, providing a structural explanation beyond simple sentiment. This framing has implications for how both companies should position their S-1 narratives.

Grok-Premium

• Broader AI company valuation landscape: Grok uniquely provided comparative valuations for other AI players — Databricks at ~$134 billion ^[12], xAI at $50–200 billion, and Perplexity at ~$20 billion ^[12] — contextualizing OpenAI and Anthropic within the broader private AI market. This matters for understanding relative valuation multiples and IPO sequencing dynamics.

• Secondary market mechanics detail: Grok specifically named "Next Round" as the platform where $600 million in OpenAI shares struggled to find buyers among "hundreds of institutions" ^[2], adding operational specificity that other providers lacked. This detail strengthens the credibility of the demand-shortage narrative.

• Anthropic's revenue growth rate context: Grok noted Anthropic's revenue was growing more than 10x annually in recent years ^[2], and that cumulative GAAP revenue through end-2025 exceeded $5 billion ^[2] — a backward-looking anchor that other providers did not surface.

OpenAI Provider

• Enterprise market share shift quantified: The OpenAI provider uniquely cited data showing OpenAI's enterprise AI API market share fell from ~50% to 25% while Anthropic's rose from 12% to 32% by end-2025 ^[29]. This is the most concrete competitive displacement metric in the entire dataset and directly explains secondary market sentiment divergence.

• Microsoft relationship as IPO risk: The OpenAI provider specifically flagged Microsoft's ~$13 billion investment and cloud services relationship as a potential IPO risk ^[2], noting OpenAI had to reaffirm that Amazon's 2026 investment does not change Microsoft ties ^[2]. This structural complexity could complicate S-1 disclosures.

• OpenAI's advertising pilot: The OpenAI provider uniquely noted that OpenAI's advertising pilot generated more than $100 million in ARR within six weeks of launch ^[16] — a revenue diversification signal not mentioned by other providers that could materially affect profitability timelines.

• OpenAI's individual investor channel: The OpenAI provider highlighted that OpenAI opened channels for individual investors for the first time, raising more than $3 billion through banks with some capital integrated into ARK Invest ETFs ^[31] — a novel democratization mechanism with implications for pre-IPO price discovery.

Perplexity

• SoftBank tranche structure: Perplexity uniquely detailed that SoftBank's $30 billion pledge is divided into three tranches arriving in April, July, and October 2026 ^[47] — meaning OpenAI's capital position is more fragile than the headline $122 billion suggests, with significant capital contingent on future milestones.

• Anthropic's customer concentration risk: Perplexity uniquely noted that approximately 25% of Anthropic's mid-2025 revenue came from just two customers ^[24], and that eight of the Fortune 10 are customers ^[24]. The concentration risk is a material IPO disclosure concern that other providers did not surface.

• Gross margin trajectory comparison: Perplexity provided the most detailed margin analysis — Anthropic targeting ~40% gross margin in 2025 (revised down due to 23% higher-than-expected cloud costs) ^[2], rising to ~77% by 2028 ^[24], versus OpenAI's estimated 30–35% ^[24]. This trajectory is central to IPO valuation modeling.

• Net Revenue Retention differential: Perplexity uniquely cited that Anthropic and Databricks both achieve ~140% NRR ^[24], while OpenAI's NRR has never been publicly disclosed and secondary research suggests it may be 110–130% ^[24] — a critical SaaS quality metric that will feature prominently in any S-1.

• Regulatory timeline: Perplexity uniquely flagged that the EU AI Act takes full effect in August 2026 ^[73] and California's SB 53 took effect January 1, 2026 ^[73] — regulatory milestones that could complicate or delay IPO filings requiring comprehensive risk disclosure.

• Prediction market probability: Perplexity cited prediction markets assigning 61% probability that either company IPOs by end of 2027 ^[43] — a useful calibration anchor for the IPO timeline uncertainty.

• HSBC cumulative loss projection: Perplexity cited HSBC analysis suggesting OpenAI could face cumulative losses of ~$115 billion through 2029 ^[2], with potential losses of ~$76 billion per annum by 2030 ^[2] — a far more bearish scenario than OpenAI's own internal projections.

Contradictions and Disagreements

Contradiction 1: OpenAI's Total Funding Round Size — $110B vs. $120B vs. $122B

Providers in conflict: OpenAI provider (citing ^[5]) vs. Grok (citing ^[2]) vs. Perplexity (citing ^[50])

The OpenAI provider noted an initial announcement of $110 billion from strategic partners ^[5], with a later close at $122 billion ^[50]. Perplexity mentioned the round was "technically extended to $120 billion" with additional sovereign wealth fund participation ^[50]. Grok cited "roughly $122 billion" ^[2]. The discrepancy likely reflects a staged close with multiple tranches, but the exact final committed figure remains ambiguous. Readers should note that SoftBank's $30 billion is structured in three tranches through October 2026 ^[47], meaning not all capital has been received.

Contradiction 2: OpenAI's Infrastructure Spend Commitment — $200B vs. $600B vs. $1.4T

Providers in conflict: OpenAI provider vs. Perplexity

The OpenAI provider cited $200 billion in infrastructure spend through 2030 ^[2]. Perplexity cited $600 billion as the current figure ^[56] and noted earlier projections were as high as $1.4 trillion before being revised down ^[56]. These figures may refer to different scopes (internal capex vs. total ecosystem commitments including Stargate), but the discrepancy is significant for cash burn modeling. This is a material unresolved contradiction.

Contradiction 3: Anthropic's Cloud Spend Commitment — $30B vs. $80B

Providers in conflict: OpenAI provider vs. Perplexity

The OpenAI provider cited Anthropic's commitment to spend $30 billion on cloud compute from Microsoft over several years ^[2]. Perplexity cited $80 billion committed with Amazon, Google, and Microsoft through 2029 ^[89], structured as operating expense. These may refer to different time periods or different subsets of cloud relationships, but the discrepancy is large enough to materially affect burn rate analysis.

Contradiction 4: OpenAI's Pre-Round Valuation — $730B vs. $840B

Providers in conflict: OpenAI provider vs. Tom's Hardware (cited by OpenAI provider)

The OpenAI provider stated OpenAI had an $840 billion valuation before the late-March 2026 round ^[5], while a Tom's Hardware source cited by the OpenAI provider referenced a $730 billion valuation after an earlier $110 billion raise ^[32]. This likely reflects sequential funding tranches at different valuations, but the sequencing is unclear.

Contradiction 5: Anthropic's Gross Margin Outlook — Revised Down vs. Trajectory Intact

Providers in conflict: Perplexity vs. Grok/Gemini

Perplexity specifically noted Anthropic revised its gross margin outlook downward due to cloud infrastructure costs rising 23% more than anticipated ^[2], with current margins around 40%. However, Grok and Gemini framed Anthropic's financial position more positively, emphasizing revenue growth without flagging the margin revision. The margin revision is a material bearish signal that Grok and Gemini underweighted.

Contradiction 6: OpenAI's Cumulative Loss Projection — $44B vs. $115B

Providers in conflict: OpenAI provider vs. Perplexity

The OpenAI provider cited OpenAI's own internal forecast of ~$44 billion in cumulative losses through 2028 ^[2]. Perplexity cited HSBC analysis suggesting ~$115 billion in cumulative losses through 2029 ^[2]. The gap between management projections and independent analyst estimates is ~$70 billion — a credibility question that will be central to any IPO roadshow.

Contradiction 7: Whether Anthropic Will IPO Before OpenAI

Providers in conflict: Grok-Premium vs. others

Grok stated Anthropic "appears more likely to IPO first" based on banker discussions, legal prep, and market sentiment ^[17]. The OpenAI provider similarly noted "many on Wall Street speculate Anthropic will go public before OpenAI" ^[2]. However, Gemini-Lite noted OpenAI CEO Altman has "pushed for a late 2026 IPO debut to stay ahead of competitors" ^[3], implying OpenAI could move faster if internal disagreements are resolved. The sequencing remains genuinely uncertain.

Detailed Synthesis

The Valuation Landscape: Record Highs, Diverging Sentiment

As of April 6, 2026, the two most prominent private AI companies in the world sit at extraordinary valuations that would have seemed implausible even 18 months ago. OpenAI closed what all four research providers confirm was the largest private tech funding round in history — approximately $122 billion at an $852 billion post-money valuation ^[50] [Perplexity, Grok, OpenAI, Gemini]. The round included commitments from Amazon ($50 billion, partially conditional on milestones), Nvidia ($30 billion), and SoftBank ($30 billion in three tranches through October 2026) ^[47] [Perplexity]. Anthropic, meanwhile, closed a $30 billion Series G in February 2026 at $380 billion ^[2] [all providers], led by GIC and Coatue, representing a more than doubling from its September 2025 valuation of $183 billion ^[58] [Perplexity].

These headline figures, however, mask a more nuanced and arguably more revealing story playing out in secondary markets.

Secondary Markets: The Real-Time Verdict

Secondary markets — where existing shareholders sell pre-IPO equity to accredited investors via platforms like Hiive, Next Round, and Augment — are providing a strikingly different signal than primary rounds [Grok]. These markets are opaque and low-volume, with pricing varying by platform and deal size ^[2] [Grok], but the directional signal in early April 2026 is unambiguous.

For OpenAI, demand has cooled sharply. Approximately $600 million in employee shares were offered on secondary platforms, and the platform Next Round struggled to find buyers among hundreds of institutions approached ^[2] [Grok, Perplexity]. The few bids that materialized implied a valuation of approximately $765 billion — roughly a 10% discount to the $852 billion primary round ^[3] [all providers]. Goldman Sachs reportedly waived its standard fees to attract buyers for OpenAI shares ^[2] [Grok, OpenAI], a remarkable concession that underscores the demand problem.

For Anthropic, the picture is inverted. Over $2 billion in buy orders flooded secondary platforms in a single week ^[2] [OpenAI], with Hiive alone logging $1.6 billion in demand ^[2] [Perplexity, OpenAI]. Bids on platforms like Next Round and Augment valued Anthropic at roughly $600 billion ^[2] [OpenAI, Grok] — a 50%+ premium to the $380 billion primary valuation from just weeks earlier ^[2] [Perplexity]. Goldman Sachs continued charging its standard 15–20% carry for Anthropic access ^[2] [Perplexity], a stark contrast to the fee waivers offered for OpenAI.

It is critical to note, as both Gemini-Lite and Grok confirmed, that no formal down round has occurred for either company ^[8] [Gemini, Grok]. The secondary discount for OpenAI reflects liquidity dynamics, sentiment shifts, and the inherent opacity of these markets — not a restructured primary valuation. Both companies also technically restrict secondary transfers without company approval and maintain rights of first refusal ^[66] [Perplexity], meaning secondary pricing is indicative rather than definitive.

The Revenue Race: Convergence and Competitive Displacement

The secondary market divergence is grounded in fundamental competitive dynamics. OpenAI reports approximately $2 billion in monthly revenue as of April 2026, equating to a ~$24 billion annual run-rate ^[2] [OpenAI, Perplexity], growing roughly 4x in the past year ^[2] [OpenAI]. Anthropic's ARR surged from approximately $14 billion in February 2026 ^[2] [OpenAI, Perplexity] to approximately $19 billion by early March 2026 ^[2] [OpenAI, Perplexity] — a pace of growth that some analyses suggest could see Anthropic surpass OpenAI in annualized revenue by mid-2026 ^[54] [Perplexity].

The most striking competitive metric, surfaced uniquely by the OpenAI provider, is the enterprise API market share shift: OpenAI's share fell from approximately 50% to 25% while Anthropic's rose from 12% to 32% by end-2025 ^[29] [OpenAI]. This is not a marginal shift — it represents a near-halving of OpenAI's enterprise dominance in roughly 12 months. Claude Code has been a primary driver, achieving a $2.5 billion+ run-rate ^[2] [Perplexity, Grok], and Anthropic now counts eight of the Fortune 10 as customers ^[24] [Perplexity].

OpenAI's revenue base remains larger and more diversified — approximately 800–900 million weekly active users ^[38] [Perplexity] and an advertising pilot that generated $100 million+ in ARR within six weeks ^[16] [OpenAI]. However, Perplexity's unique finding that approximately 25% of Anthropic's mid-2025 revenue came from just two customers ^[24] introduces a customer concentration risk that will require prominent S-1 disclosure.

The Cash Burn Problem: Existential for Both, Worse for OpenAI

Both companies face severe cash burn, but the scale and trajectory differ materially. OpenAI's internal forecasts project approximately $14 billion in net losses for 2026 alone ^[2] [Grok, OpenAI, Perplexity], with cumulative losses through 2028 projected at ~$44 billion by management ^[2] [OpenAI] — though HSBC's independent analysis cited by Perplexity suggests cumulative losses could reach ~$115 billion through 2029 ^[2] [Perplexity], with potential annual losses of $76 billion by 2030. The gap between management and analyst projections ($70 billion) is a material credibility issue for any IPO roadshow.

OpenAI has committed to spend $200 billion on infrastructure through 2030 per the OpenAI provider ^[2], though Perplexity cites $600 billion as the current figure with earlier projections as high as $1.4 trillion ^[56] — a significant unresolved discrepancy that likely reflects different scopes of commitment. Some analysts have warned OpenAI could run out of cash by mid-2027 without continued capital raises ^[41] [OpenAI]. The SoftBank tranche structure — $30 billion spread across April, July, and October 2026 ^[47] [Perplexity] — means OpenAI's capital position is more fragile than the $122 billion headline suggests.

Anthropic's burn profile is less severe but still substantial. The company has committed approximately $80 billion with cloud providers Amazon, Google, and Microsoft through 2029 ^[89] [Perplexity], structured as operating expense. Gross margins were revised downward after cloud infrastructure costs rose 23% more than anticipated ^[2] [Perplexity], with current margins around 40% — though the trajectory toward ~77% by 2028 ^[24] [Perplexity] is more credible than OpenAI's margin path. Anthropic's net revenue retention of ~140% ^[24] [Perplexity], compared to OpenAI's estimated 110–130% ^[24] [Perplexity], suggests higher-quality enterprise revenue.

IPO Timelines: Contested, Uncertain, Consequential

The IPO question is where the most significant uncertainty — and the most consequential near-term decisions — reside. Prediction markets as of early April 2026 assigned only a 61% probability that either company completes an IPO by end of 2027 ^[43] [Perplexity], reflecting genuine uncertainty despite the intense speculation.

Anthropic is the more likely near-term candidate by most assessments [Grok, OpenAI]. The company has engaged Wilson Sonsini as legal counsel ^[2] [Grok], is in active discussions with Goldman Sachs, JPMorgan, and Morgan Stanley ^[2] [Grok, Gemini], and leadership has discussed a Q4 2026 target ^[3] [all providers]. Bankers have floated a potential raise exceeding $60 billion ^[3] [Grok, OpenAI, Perplexity]. The secondary market enthusiasm at ~$600 billion provides a favorable backdrop, though the gap between secondary bids and banker estimates of $400–500 billion [Gemini] suggests some froth in secondary pricing.

OpenAI's IPO path is complicated by the CEO-CFO divergence confirmed by all four providers. Sam Altman targets Q4 2026 ^[3] [all providers], motivated by competitive pressure and momentum. CFO Sarah Friar has raised concerns about organizational readiness, procedural gaps, and the sustainability of aggressive spending plans ^[2] [Gemini, Grok, OpenAI, Perplexity], advocating for a 2027 timeline. OpenAI's structural complexity — a Public Benefit Corporation operating under nonprofit oversight ^[77] [OpenAI, Perplexity], a complicated Microsoft relationship involving ~$13 billion in investment and cloud services ^[2] [OpenAI], and the need to disclose $14 billion in projected 2026 losses — creates substantial S-1 preparation challenges.

The broader IPO market context is mixed: the Federal Reserve has signaled only one 25 basis point rate cut in 2026, with rates remaining at 3.5–3.75% ^[87] [Perplexity], and the EU AI Act takes full effect in August 2026 ^[73] [Perplexity], adding regulatory disclosure complexity. Databricks, a useful comparable at $134 billion valuation with positive free cash flow and 80%+ subscription gross margins ^[79] [Perplexity, Grok], may provide a more favorable IPO template than either OpenAI or Anthropic can currently match.

Valuation Framework: What the Numbers Actually Mean

The enterprise LLM market is projected to grow from $5.91 billion in 2026 to $48.25 billion by 2034 at a 30% CAGR ^[61] [Perplexity], providing the fundamental demand backdrop. Against this, OpenAI's $852 billion primary valuation implies a revenue multiple of approximately 35x current ARR — extraordinary even by AI standards. Anthropic's $380 billion primary implies approximately 20–27x ARR depending on which revenue figure is used, while secondary market bids at $600 billion imply ~31x — suggesting secondary buyers are pricing in continued hypergrowth.

The key valuation risk for both companies, as Grok noted ^[18], is that primary rounds often include strategic upside or illiquidity premiums that secondary markets — which are more liquidity-sensitive and real-time — do not sustain. The divergence between OpenAI's primary and secondary pricing, and between Anthropic's primary and secondary pricing (in opposite directions), illustrates this dynamic vividly.