Fireworks: DeepSeek-V4.1-Flash Matches DeepSWE Quality at 1/15th the Cost
lqiao · x · 2026-09-16
Fireworks AI benchmarked its new DeepSeek-V4.1-Flash, claiming a new Pareto frontier on quality vs. cost. Key insight: running DeepSWE, input tokens outnumbered output 174:1, 99.6% were cache hits, and those hits made up 60% of the bill — netting $0.43/task vs $6.52 at equal quality. Agent cost optimization hinges on input caching, not output pricing.
More from Infra
- Palmer Luckey asks: can you use datacenter RAM for things prohibited by the operator's terms of service? — avaitopiper · 2026-11-02
- Disable MTP for 20-30% faster local LLM inference on limited RAM — InternalMode8159 · 2026-11-02
- Kioxia up 216% in a year but down 68% from its June 2026 peak — zephyr_z9 · 2026-11-02
- Intel plunges 29%, worst day since the 1970s, wiping $12B off US government stake — emmanuelvivier · 2026-11-02
- Researcher hoards nearly 1TB of DRAM, refuses to sell a stick to data centers — kevinnbass · 2026-11-02
- Self-Hosting LLMs at a Law Firm: Infra and Data Governance Harder Than the Model — EducationalGroup7245 · 2026-11-02