Every major LLM's list price per 1M tokens in one table — from GPT-5 at $5.00/$15.00 to DeepSeek V3 at $0.27/$1.10. Tracked weekly, free forever.
Last verified 2026-08-16 · 25 models · 11 providersList prices in USD per 1M tokens (input / output). Image models are priced per generated image. Prices verified 2026-08-16.
| Model | Provider | Context window | Input $/1M | Output $/1M | Notes |
|---|---|---|---|---|---|
| Claude Opus 4 | Anthropic | 500K | $15.00 | $75.00 | Most capable; enterprise workloads |
| GPT-5 | OpenAI | 400K | $5.00 | $15.00 | General frontier workhorse |
| GPT-5.6 Sol | OpenAI | 400K | $4.00 | $16.00 | Flagship reasoning; deep research |
| Claude Sonnet 4 | Anthropic | 400K | $3.00 | $15.00 | Balanced coding & agentic work |
| Grok 4.1 | xAI | 256K | $1.50 | $6.00 | Reasoning + real-time knowledge |
| Gemini 2.5 Pro | 1M | $1.25 | $5.00 | 1M-token context window | |
| GPT-5.4 | OpenAI | 300K | $1.25 | $5.00 | Mid-tier reasoning |
| Mistral Large 3 | Mistral | 128K | $1.00 | $3.00 | EU enterprise; on-prem friendly |
| Llama 4 405B | Meta | 256K | $0.90 | $0.90 | Open weights; symmetric in/out |
| Claude Haiku 4 | Anthropic | 300K | $0.80 | $4.00 | Fast assistant; high volume |
| Kimi K2.6 | Moonshot AI | 256K | $0.60 | $2.50 | Open weights; long-context |
| DeepSeek R1 | DeepSeek | 128K | $0.55 | $2.19 | Open reasoning model |
| Grok Chat Fast | xAI | 256K | $0.50 | $2.00 | Consumer chat; low latency |
| o3-mini | OpenAI | 200K | $0.40 | $1.60 | Fast small reasoning |
| GLM-4 | Zhipu AI | 128K | $0.35 | $1.40 | Open weights; strong coding |
| Gemini 2.5 Flash | 1M | $0.30 | $2.50 | High-volume; near-real-time | |
| Qwen 3.5 | Alibaba | 256K | $0.30 | $1.20 | Open weights; multilingual |
| DeepSeek V3 | DeepSeek | 128K | $0.27 | $1.10 | Open weights; best value chat |
| GPT-5.4-mini | OpenAI | 300K | $0.20 | $0.80 | Low-latency mini |
| GPT-5-mini | OpenAI | 400K | $0.15 | $0.60 | Cheapest frontier mini of 2026 |
| Midjourney V7 | Midjourney | — | $5.00/img | — | 2K/4K renders |
| Nano Banana Pro | — | $4.00/img | — | 4K photoreal editing | |
| GPT Image 2 | OpenAI | — | $3.00/img | — | Premium generation tier |
| GPT Image 1.5 | OpenAI | — | $2.50/img | — | Mid-tier image generation |
| GPT Image 1 | OpenAI | — | $2.00/img | — | Entry image generation |
Ranked by input price per 1M tokens. All five cost less than a single cent per page of text.
The premium tier — where frontier capability still commands a premium. Ranked by input price per 1M tokens.
List-price history for frontier models, USD per 1M input tokens (yearly average).
| Year | Avg frontier input $/1M | Cheapest flagship input $/1M | Example |
|---|---|---|---|
| 2024 | $10.00 | $2.50 | GPT-4o at $2.50; GPT-4 Turbo at $10.00 |
| 2025 | $4.20 | $0.27 | DeepSeek V3 at $0.27 reset the market |
| 2026 | $2.10 | $0.15 | GPT-5-mini at $0.15; mini-tier goes sub-$0.20 |
Stop metering tokens. DrAI bundles GPT-5.6 Sol, Claude Opus 4, DeepSeek V3 and 20+ more models into a single subscription with a unified API.
See DrAI pricing Start free — no card needed