DeepInfra selling DeepSeek v4.1 at 49% off sparks doubts it can turn a profit
gharik · x · 2026-10-19
Greg Kamradt questions whether DeepInfra can actually make money serving DeepSeek v4.1 at rock-bottom prices. A provider comparison chart shows DeepInfra's 51%-off rate at just $0.098/M input and $0.294/M output (Flex tier: $0.0784/$0.2352), versus $1.20/M output at Fireworks and Baseten — nearly a 4x gap. Baseten leads on latency (0.28s, 229 tps), while newer players like inference.net and Relace undercut at $0.05-0.08/M input. Inference-hosting price competition is heating up.
More from Venture
- Stripe's Link now powers ~200 AI agents as agentic commerce volume turns 'material' — jeff_weinstein · 2026-10-19
- Barclays: pure-play data center bond spreads hit ~209bps as market prices in contagion — SumitGup · 2026-10-19
- A decade after Theranos, skeptics ask if AI diagnostics will be worse — krishnan · 2026-10-19
- Anthropic's $2T Valuation Tops 35 SaaS Firms Combined Despite 4.6x Less Revenue — Soggy_Perspective776 · 2026-10-19
- Polymarket puts 54% odds on Anthropic IPO in November — Polymarket · 2026-10-19
- Opinion: The AI Bubble Is About to Burst Like the Dot-Com Crash — beingmodest · 2026-10-19