Cloud vs. Rig TCO
Plot cumulative 36-month API spending against hourly RunPod/H100 rentals and owned RTX/Mac hardware with electricity and resale value.
Low-latency classification with cached JSON schemas across high-frequency SaaS events.
Multi-turn history expands average input context to 1,200 tokens/req. Output accounts for 57% of total spend.
| Model | Delta | |||
|---|---|---|---|---|
| Claude Haiku 5.5 (High-Volume)Anthropic | $0.10 | $0.50 | $158.88 | Active Baseline |
| GPT-4.1 nano (Ultra-Low Cost)OpenAI | $0.10 | $0.40 | $159.8 | +$1 |
| DeepSeek V4.1 Flash (Off-Peak)DeepSeek | $0.15 | $0.60 | $195.59 | +$37 |
| GPT-4o mini (Legacy Budget)OpenAI | $0.15 | $0.60 | $287.63 | +$129 |
| GPT-4.1 mini (Balanced Fast)OpenAI | $0.40 | $1.60 | $639.19 | +$480 |
| Gemini 3.5 Flash-Lite (Budget Edge)Google | $0.30 | $2.50 | $659.28 | +$500 |
| DeepSeek V4 Pro (Off-Peak Reasoning)DeepSeek | $0.66 | $1.98 | $751.32 | +$592 |
| Gemini 3.8 Flash (Intro Paid Tier)Google | $0.75 | $3.75 | $1,191.63 | +$1,033 |
| Llama 3.3 / 4 70B-Class (Hosted API)Open-Weight Host | $0.59 | $0.79 | $1,221.76 | +$1,063 |
| o3-mini (Reasoning)OpenAI | $1.10 | $4.40 | $2,109.32 | +$1,950 |
| Gemini 4 HydrogenGoogle | $1.50 | $7.50 | $2,287.38 | +$2,129 |
| GPT-5 (Flagship Reasoning)OpenAI | $1.25 | $10.00 | $2,670.89 | +$2,512 |
| Claude Sonnet 5.5 (Coding & Agents)Anthropic | $2.00 | $10.00 | $3,049.84 | +$2,891 |
| GPT-6.2 SolOpenAI | $2.00 | $10.00 | $3,049.84 | +$2,891 |
| Gemini 4 CarbonGoogle | $2.00 | $10.00 | $3,049.84 | +$2,891 |
| GPT-4.1 (1M Long-Context)OpenAI | $2.00 | $8.00 | $3,195.94 | +$3,037 |
| Gemini 3.1 Pro (Long-Context)Google | $2.00 | $12.00 | $3,542.93 | +$3,384 |
| Gemini 4 NeonGoogle | $3.00 | $15.00 | $4,574.76 | +$4,416 |
| GPT-4o (Legacy Omni)OpenAI | $2.50 | $10.00 | $4,793.91 | +$4,635 |
| Claude Opus 5.5 (Agentic Frontier)Anthropic | $4.00 | $20.00 | $6,099.68 | +$5,941 |
| Claude Fable 5.2Anthropic | $8.00 | $40.00 | $11,943.68 | +$11,785 |
| GPT-6.2 AstraOpenAI | $8.00 | $40.00 | $12,199.35 | +$12,040 |
High cache hit on identical taxonomy schemas dominates efficiency. Pre-populated at 50,000 daily requests with 1,200 input tokens, 120 output tokens, 70% prompt caching, and 1 turns/session.
Primary Technical & US Grid Benchmarks
Plot cumulative 36-month API spending against hourly RunPod/H100 rentals and owned RTX/Mac hardware with electricity and resale value.
Count live prompt tokens across English prose and JSON/code, size RAG document chunks, and compute float32 vs. int8 HNSW vector RAM.
Test Llama 70B, Qwen 32B, or DeepSeek R1 across RTX 3090/4090/5090 & Apple Silicon Macs with GQA KV-cache math and tokens/sec speed.