Observed Signal · Jul 7, 2026 · Analysis · Source: techcrunch · Impact: 3/5 · Sentiment: Neutral
Open‑Source Models Haven't Hurt Anthropic Yet
TechCrunch analysis argues that the rise of open‑source AI models has not materially reduced spending on frontier models from firms like Anthropic. Decagon CEO Jesse Zhang's theory suggests frontier models are used for discovery and proof-of-concept work while mature production use cases migrate to cheaper open‑source alternatives. Data from dashboards such as Vercel’s AI gateway and OpenRouter show open‑source models (e.g., DeepSeek V4 Flash) dominating token volumes, while Anthropic still captures a majority of token spend on some platforms because premium frontier models carry much higher per‑token prices. Nvidia’s Nemotron is cited as a new entrant likely to shift rankings. The piece frames a two‑tier model economy — frontier labs owning discovery and open source owning production — as a stable feature of the current AI market.
The piece analyzes the economic relationship between expensive frontier models and lower‑cost open‑source models, citing token volumes and spend metrics that influence vendor choice and AI infrastructure budgeting across enterprises — relevant to companies procuring LLM inference for MarTech/AdTech use cases.
Track Anthropic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Decagon CEO Jesse Zhang published a post titled “Everyone is wrong about open source AI in the enterprise.”
- Vercel’s AI gateway dashboard showed DeepSeek surged into the lead for token volumes, processing just over one third of tokens on the platform in the referenced week.
- Anthropic accounts for more than half of overall AI spend on Vercel’s platform despite open‑source models growing in usage.
- OpenRouter reported DeepSeek V4 Flash processing 5.3 trillion tokens weekly while Opus 4.8 handled just over 2 trillion tokens; OpenRouter registered Opus 4.8’s average token cost at roughly $1.37 per million tokens versus V4 Flash at about $0.06 per million (≈23x difference).
- Nvidia’s new model Nemotron is described as poised to leap to the front of the pack due to Nvidia’s relationships and the model’s adaptability.
Connected Companies & Entities
6 Entities mapped“But if you scroll down to overall token spend, you’ll see Anthropic still accounts for more than half of the overall AI spend on the platfor...”
“Vercel’s AI gateway dashboard shows that, in just the past week, DeepSeek has surged into the lead for token volumes, now processing just ov...”
“Vercel’s AI gateway dashboard shows that, in just the past week, DeepSeek has surged into the lead for token volumes, now processing just ov...”
“Z.ai — the lab behind the popular GLM-5.2 model — jumped into a respectable fourth place over the same period....”
“OpenRouter tells a similar story, capturing a much larger (but slightly less enterprise-y) segment of the market....”
“Those figures don’t even capture the newest arrival, Nvidia’s Nemotron, which is poised to leap to the front of the pack by virtue of Nvidia...”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Rising AI Costs Renew Interest in Open-Source Models
Rising inference costs and geopolitical frictions have revived industry interest in open-source AI models as a lower-cost, more controllable alternative to frontier proprietary systems. The article reports that Chinese labs have pulled ahead since DeepSeek’s 2025 breakout, with open models such as Qwen and Kimi becoming common foundations for startups. Anthropic’s Mythos family faces export-control scrutiny, adding urgency to alternatives. Poolside — a $12 billion model maker backed by Nvidia, Bain Capital and DST — has publicly refocused on open source and released Laguna XS.2 in April, a model tailored for agentic coding, according to co‑founder and co‑CEO Eiso Kant. Investors including Michael Mignano (USV) and Bill Gurley say startups and companies are rethinking costly “tokenmaxxing” and diversifying model usage to control spend.
Open models challenge frontier AI's centrality
TechCrunch reports that open-weight AI models and open-source ecosystems are gaining substantial production share, challenging the primacy of proprietary 'frontier' models. Chinese open-weight models made up 41% of Hugging Face downloads this spring and dominate popularity rankings on OpenRouter. Platforms such as Vercel show open models handling roughly a third of AI requests in June, while Hugging Face says it hosts millions of public models and datasets and sees rapid repository growth. Executives including Hugging Face CEO Clem Delangue and Microsoft CEO Satya Nadella argue enterprises prefer ownership and control over rented black-box models, while Anthropic CEO Dario Amodei warns about dangers from widely released powerful weights. The piece frames the shift as a trade-off between decentralization/transparency and risks tied to broad availability of capable models.
OpenAI, Anthropic Face Shift From Tokenmaxxing to Efficiency
Enterprise customers are reining in runaway AI token spending and shifting from 'tokenmaxxing' toward cost-efficient model use, putting pressure on leading model providers OpenAI and Anthropic. Startups and enterprises are routing tasks to cheaper models, switching providers (one startup moved all traffic from Anthropic to Chinese firm DeepSeek), and implementing usage caps and analytics to control bills. The shift comes as both Anthropic and OpenAI report multibillion-dollar annualized run rates and weigh IPO timing; investors and analysts note urgency to list before corporate customers rationalize AI spend. Big cloud and platform vendors — Microsoft, Amazon and Google — are promoting lower-cost alternatives and model-routing features, intensifying competition. Vendors have added enterprise spend controls and analytics, while finance leaders and consultants urge proving ROI before large-scale deployments.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.