AI API Price Tracker
Business Task Calculator ALL PRICING IN USD ($)
Estimate API costs for real-world tasks. (Assumes 500 words per task, 3:1 Input/Output ratio).
Telemetry Definitions
Composite benchmark score. >80 indicates reliability for autonomous agentic workflows.
Estimated USD cost per 1M tokens via standard 3:1 ratio (In/Out) to prevent hidden output costs.
TPS equals generation speed. TTFT dictates latency (crucial for Voice UIs).
Formula: (IQ² / (Cost + 1)) / 100. Exponentially rewards intelligence while penalizing overhead.
AI model API pricing and benchmarks, updated August 12, 2026
The AI Profit Wire tracks 375 language models on blended cost per million tokens at a 3:1 input to output ratio, intelligence index, and output speed. The highest scoring model tracked is Claude Opus 5 by Anthropic, at $10.00 per million tokens. The lowest cost model tracked is Gemma 4 E4B by Google, at $0.040 per million tokens.
| Model | Provider | Blended price (USD per 1M tokens) | Intelligence index | Output speed (tokens/sec) |
|---|---|---|---|---|
| Claude Opus 5 | Anthropic | $10.00 | 63.1 | 53 |
| Claude Fable 5 | Anthropic | $20.00 | 62.1 | 61 |
| GPT-5.6 Sol | OpenAI | $11.25 | 60.9 | 68 |
| Kimi K3 | Kimi | $6.00 | 59.7 | 37 |
| Qwen3.8 Max | Alibaba | $3.00 | 58.1 | 69 |
| Claude Opus 4.8 | Anthropic | $10.00 | 57.3 | Not measured |
| Muse Spark 1.2 | Meta | $2.00 | 56.8 | Not measured |
| GPT-5.6 Terra | OpenAI | $4.50 | 56.6 | 125 |
| GPT-5.5 | OpenAI | $11.25 | 56.3 | Not measured |
| Grok 4.5 | SpaceXAI | $3.00 | 55.8 | 53 |
| Claude Sonnet 5 | Anthropic | $4.00 | 55.3 | 78 |
| Claude Opus 4.7 | Anthropic | $10.00 | 55 | Not measured |
| Muse Spark 1.1 | Meta | $2.00 | 53.2 | 192 |
| GPT-5.4 | OpenAI | $5.63 | 53.1 | Not measured |
| GLM-5.2 | Z AI | $2.02 | 52.6 | 115 |
| GPT-5.6 Luna | OpenAI | $0.450 | 52.3 | 171 |
| Gemini 3.5 Flash | $3.38 | 52 | 226 | |
| DeepSeek V4 Flash 0731 | DeepSeek | $0.175 | 51.8 | 103 |
| Gemini 3.6 Flash | $3.00 | 51.6 | 210 | |
| Claude Sonnet 4.6 | Anthropic | $6.00 | 48.4 | Not measured |
| Gemini 3.1 Pro Preview | $4.50 | 47.7 | 128 | |
| Qwen3.7 Max | Alibaba | $3.75 | 46.7 | 195 |
| GPT-5.3 Codex | OpenAI | $4.81 | 45.5 | 99 |
| MiniMax-M3 | MiniMax | $0.525 | 45.4 | 81 |
| DeepSeek V4 Pro | DeepSeek | $0.544 | 45.3 | 59 |
| Kimi K2.6 | Kimi | $1.71 | 45.1 | Not measured |
| Claude Opus 4.6 | Anthropic | $10.00 | 44.9 | Not measured |
| GPT-5.2 | OpenAI | $4.81 | 43.3 | Not measured |
| Kimi K2.7 Code | Kimi | $1.71 | 43 | 40 |
| MiMo-V2.5-Pro | Xiaomi | $0.544 | 42.9 | 72 |
| Inkling | Thinking Machines | $1.76 | 42.3 | 77 |
| Hy3 | Tencent | $0.241 | 42.2 | 66 |
| Nex-N2-Pro | Nex AGI | $1.00 | 42.1 | 124 |
| DeepSeek V4 Flash | DeepSeek | $0.168 | 42.1 | Not measured |
| Claude Opus 4.5 | Anthropic | $10.00 | 41.9 | Not measured |
| Inkling Small | Thinking Machines | $0.525 | 41.2 | 119 |
| GPT-5.2 Codex | OpenAI | $4.81 | 41.2 | Not measured |
| Qwen3.6 Max Preview | Alibaba | $2.93 | 41.1 | Not measured |
| GLM-5.1 | Z AI | $2.14 | 41 | Not measured |
| GPT-5.4 mini | OpenAI | $1.69 | 40.9 | Not measured |
| Grok Build 0.1 0616 | SpaceXAI | $1.25 | 40.7 | Not measured |
| Gemini 3 Pro Preview | $4.50 | 40.6 | Not measured | |
| GLM-5 | Z AI | $1.55 | 40.6 | Not measured |
| Qwen3.6 Plus | Alibaba | $1.13 | 40.5 | 55 |
| Agnes 2.5 Pro Alpha | Sapiens AI | $0.563 | 39.7 | 130 |
| GPT-5.4 nano | OpenAI | $0.463 | 39.7 | Not measured |
| Qwen3.7 Plus | Alibaba | $0.700 | 39.4 | 55 |
| MiniMax-M2.7 | MiniMax | $0.525 | 38.9 | Not measured |
| Gemini 3 Flash Preview | $1.13 | 38.7 | Not measured | |
| Nemotron 3 Ultra 550B A55B | NVIDIA | $1.14 | 38.3 | 139 |