Who Actually Pays Together, Fireworks and DeepInfra? A Reddit Debate
Azamat_Kuzdibay · reddit · 2026-09-12
A Reddit thread questions the business fundamentals of open-model inference providers. The poster observes that most teams just call OpenAI/Anthropic/Gemini APIs, and when cost becomes an issue they downgrade to cheaper closed models like Gemini Flash rather than switching to open weights. Commenters point to the real customer base: enterprises needing data compliance, private deployment, stable high-volume throughput, and hosting of fine-tuned open models — use cases closed APIs can't serve.
More from Venture
- SoftBank's Masayoshi Son Predicts 100 Trillion Self-Replicating AIs That Will Surpass Humans — Puzzleheaded-King584 · 2026-09-12
- Coinbase Data Shows Perplexity Leads Agentic Trading Volume at 26.8% — AravSrinivas · 2026-09-12
- Open-source models account for just ~3.2% of global AI training spend, analysis finds — 0xBekket · 2026-09-12
- Boring Company's Steve Davis: tunnels at $3-5M per mile make epic projects possible — elonmusk · 2026-09-12
- Nscale adds former OpenAI exec Fidji Simo to board ahead of potential IPO — TechCrunch AI · 2026-09-12
- CPO Doesn't Kill Optics, It Moves the Rent: Optical Coupling Is the Real Throughput Bottleneck — demian_ai · 2026-09-12