Anthropic · seit 15. September 2026 (Early Access) / September 15, 2026 (early access) · 2× · zuletzt 06. Okt. 2026
WICHTIGER HINWEIS: "Jev" ist entgegen der Aufgabenstellung KEIN Anthropic-Produkt. Laut offizieller Produktseite (typesafe.ai, docs.typesafe.ai) und übereinstimmenden Tech-Medien wurde Jev von TypeSafe AI entwickelt, einem 2024 gegründeten Startup von Diogo Almeida (ex-OpenAI), Erik Gafni und Sasha Sheng. Anthropic (Claude) wird in Berichten lediglich als Wettbewerber bzw. als Hilfsmodell in einem unabhängigen Drittprojekt (Pokémon-Experiment) erwähnt, ist aber nicht Hersteller oder Betreiber von Jev. Jev ist ein sogenanntes "System One"-Modell: Es generiert keinen Fließtext, sondern liefert ausschließlich typisierte, kalibrierte Entscheidungen (Choice, Score, Noul) mit Wahrscheinlichkeiten in einem einzigen Inferenzdurchgang. Es ist primär für Klassifikations-, Routing- und Scoring-Aufgab
Features
| Durchsatz/Latenz | End-to-End-Latenz ca. 70–500 ms; Ratenlimits ca. 100.000 Tokens/Sek. und 80 Requests/Sek. laut offizieller Doku |
| Lizenz | Proprietär (Closed-Source, nur API-Zugang) |
| Plattform | Eigene API (api.typesafe.ai), zusätzlich über OpenRouter, Vercel AI Gateway und Cloudflare Workers AI verfügbar |
| Preis | $0,042 pro 1 Mio. Input-Tokens; Output-Tokens kostenlos |
| Protokoll-Kompatibilität | Eigenes Protokoll über einen einzigen Endpunkt POST /v1/systemone; kein OpenAI- oder Anthropic-kompatibles Format |
| Release-Datum | 15. September 2026 (Early Access, Warteliste) |
| Unterstützte Modelle/Provider | Nur Jev selbst (Modell-Aliasse jev-latest, jev-1.13.0); separates Python-Adapter-Tool unterstützt Vergleich mit OpenAI-, Anthropic- und Gemini-Modellen |
Weitere Produkte in dieser Kategorie: LLM-Inferenz & Serving
Belege (2)
Unternehmens-Analyse: Anthropic
Stand 22.8.2026Anthropic combines exceptional revenue growth, strong product‑market fit, and powerful distribution via major cloud partners, making it one of the most strategically important AI companies globally. However, the latest funding rounds imply a ~$380B post‑money valuation against even the most aggressive reported revenue run‑rate figures, resulting in a revenue multiple that is extremely demanding relative to both hyperscalers and leading AI peers. With competition intensifying, capital requirements remaining enormous, and public‑market investors likely to scrutinize profitability and unit economics more harshly than private capital, the balance of risk and reward at current levels appears skewed to the downside over a 2–3 year horizon. While long‑term optimists may still view Anthropic as a potential AI platform giant, from a valuation‑sensitive equity perspective the prudent stance at today’s implied pricing is SELL (or underweight), with a preference to re‑enter at lower multiples or after clearer evidence of sustainable, high‑margin economics post‑IPO.
Key Takeaways
- Anthropic has scaled from roughly $1B annualized revenue at the end of 2024 to over $5B run-rate by August 2025 and, according to more recent disclosures, above $30B and then $65B annualized revenue by mid‑2026, implying triple‑digit growth and one of the fastest enterprise software ramps in history.
- Valuation has expanded even faster than revenue: from ~$18.4B post‑money in early 2024 to $61.5B (Series E, March 2025), $183B (Series F, September 2025) and ~$380B post‑money in a $30B Series G round in early 2026, implying a current revenue multiple that is extremely rich even versus top‑tier hyperscalers and AI peers.
- Product strength is anchored in the Claude 3/3.5 and subsequent model families, which independent evaluations (e.g., NIST AISI) and benchmarks show as competitive or superior to OpenAI and others on many reasoning and coding tasks, and deeply integrated into Amazon Bedrock and Google Vertex AI, supporting durable enterprise distribution.
- Despite rapid revenue growth and strong strategic partnerships, Anthropic faces intense competition (OpenAI, Google, Meta, others), massive capital intensity for compute, and execution risk around safety, regulation, and eventual IPO pricing, all of which could compress valuation multiples from current levels.
- Google Trends data over the last two years shows global search interest in “Anthropic” rising sharply from mid‑2024 through late‑2025, peaking around the large funding/valuation announcements in late‑2025 to early‑2026, then stabilizing at a high plateau rather than declining, suggesting sustained but no longer accelerating public attention (RISING → STABLE at elevated level).
Action-Ideen
For late‑stage private investors or secondary‑market buyers at or near the ~$380B post‑money valuation, the risk‑reward skews negative: even assuming $65B+ revenue run‑rate and continued hypergrowth, the implied forward revenue multiple is well above large‑cap software and AI leaders, leaving little room for multiple expansion and significant downside if growth normalizes or competition intensifies.
Horizont: 24 Mon.
For very long‑term, high‑risk‑tolerant investors able to access primary rounds or structured exposure, Anthropic offers a rare chance to own a leading foundation‑model provider with best‑in‑class growth, strong enterprise traction via AWS/Google partnerships, and a credible path to becoming a core layer of the AI stack; if it can sustain leadership and scale to hyperscaler‑like revenue, current valuation could still be justified over a 5–7 year horizon.
Horizont: 72 Mon.
Existing investors from earlier rounds (e.g., pre‑2025) sit on very large unrealized gains; given Anthropic’s exceptional growth but stretched valuation, a balanced approach is to hold core exposure while gradually taking liquidity on strength or around IPO, recognizing both the upside optionality from continued dominance and the non‑trivial risk of multiple compression.
Horizont: 36 Mon.
Google Trends · ↗ steigend
Global Google search interest for “Anthropic” over the last two years was relatively modest through early‑2024, then began to climb with the Claude 3.5 Sonnet launch and major partnership news, accelerated sharply around the large funding and valuation rounds in 2025, and has since stabilized at a high plateau; the overall two‑year pattern is a strong uptrend that has recently transitioned from steeply rising to broadly stable at elevated levels rather than rolling over.
Contrarian Insights
- • While consensus focuses on Anthropic’s headline valuation as evidence of a bubble, a contrarian view is that if the company can translate its current $30–65B+ revenue run‑rate into a durable, high‑margin software and platform business at hyperscaler scale, today’s multiples could compress naturally over time without destroying equity value, making the current price less extreme on a 7–10 year view than it appears on a 1–3 year view.
- • Many investors assume OpenAI and the largest hyperscalers will inevitably dominate foundation models, but Anthropic’s strong safety brand, close work with regulators (e.g., AISI evaluations), and deep integration into multiple clouds position it as a neutral, multi‑cloud alternative; in a world where enterprises and governments seek diversification and safety‑first providers, Anthropic could capture a structurally larger share of high‑value, regulated workloads than the market currently discounts.
Quellen (8)
- https://www.anthropic.com/news/anthropic-raises-30-billion-series-g-funding-380-billion-post-money-valuation
- https://www.anthropic.com/news/anthropic-raises-series-f-at-usd183b-post-money-valuation
- https://www.anthropic.com/news/anthropic-raises-series-e-at-usd61-5b-post-money-valuation
- https://stockanalysis.com/private/anthropic/valuation/
- https://www.anthropic.com/news/claude-3-5-sonnet
- https://www.nist.gov/news-events/news/2024/11/pre-deployment-evaluation-anthropics-upgraded-claude-35-sonnet
- https://www.visualcapitalist.com/charted-the-soaring-revenues-of-ai-companies-2023-2025/
- https://www.whitehouse.gov/wp-content/uploads/2025/03/Artificial-Intelligence-and-the-Great-Divergence.pdf