The most intelligence for the money, right now
Every model below is one no other model beats on both counts: none scores higher for less. The blended price is a million tokens at three input to one output, the usual mix of a chat or an agent; cached input is what a repeated prompt prefix costs instead. Every figure is a record re-read every 6 hours, so the table changes when a lab reprices.
| Model | Lab | Index | Blended / 1M | Input / 1M | Output / 1M | Cached input / 1M |
|---|---|---|---|---|---|---|
| Nemotron 3.5 Lightning | NVIDIA | 5.5 | $0.1025 | $0.07 | $0.20 | $0.04 |
| MiMo-V2.6-Flash | Xiaomi | 26.7 | $0.175 | $0.14 | $0.28 | $0.0028 |
| Claude Haiku 5.5 | Anthropic | 36.2 | $0.20 | $0.10 | $0.50 | $0.01 |
| GPT-6.1 Terra | OpenAI | 51 | $2 | $1 | $5 | $0.05 |
| GPT-6.1 Sol | OpenAI | 54.5 | $4 | $2 | $10 | $0.10 |
| Claude Opus 5.5 | Anthropic | 58.4 | $8 | $4 | $20 | $0.20 |
Every model, its price and its window
Every guide
- AI Receptionist Cost per Call: Build vs Buy (2026): AI receptionists answer calls for ~8 cents vs $10+ for humans. Compare build, platform, and buy costs per call to choose the right option in 2026.
- Claude Max API Credits: Claim Them, Ship an App (2026): Claim your Claude Max or Team API credit into a capped workspace and build an app on Haiku 5.5 before the monthly credit expires unused.
- Sign in with ChatGPT API: Users Pay Your AI Bill (2026): Learn how Sign in with ChatGPT lets users' own plans pay for your app's AI costs, what limits apply, and how to build it safely in 2026.
- Cut AI Costs With Prompt Caching (2026): Reduce AI costs by 3-5x with prompt caching. Master implementation for Anthropic, OpenAI, Gemini, and DeepSeek to optimize your 2026 AI budget.
- GPT-6 Astra Pricing: What Your App Really Pays 2026: Master GPT-6 Astra's 272K pricing cliff. Learn token efficiency strategies to cut costs and build affordable agentic AI features in 2026.
- Set Up Metered Billing for Your AI Product 2026: Set up metered billing for your AI product in 2026. Compare Stripe, Metronome, Orb, Lago and merchant-of-record options, with real pricing, code, and the traps.
- Price Your AI Product to Beat Token Costs 2026: Master AI pricing strategies to protect margins and scale effectively as token costs evolve. Learn hybrid models for sustainable growth.
- Set the Effort Dial to Cut AI Costs (2026): Learn how to slash AI expenses in 2026 by adjusting model reasoning effort to prevent overthinking on simple tasks.
- Cut AI Agent Costs With Model Routing (2026): Learn how to slash AI agent costs by routing each request to the cheapest capable model. Includes playbooks, platforms, and failure modes to avoid.
- What It Costs to Build an App With AI 2026: Discover the real 2026 cost to build an AI app: token meters, hidden bills, and honest budgets before you spend a cent.