Prices verified October 3, 2026
What will your AI API actually cost?
Compare 24 models from OpenAI, Anthropic, Google, xAI and DeepSeek side by side — prices, benchmarks and real-workload estimates instead of per-token guesswork.
Already have a prompt? Count its tokens and cost →
Popular comparisons
All comparisons →Capability vs. price
Full benchmarks →Most capable
- GPT-6 AstraECI 166.3 · $20.00
- Claude Fable 5.1ECI 164.5 · $20.00
- Claude Opus 5ECI 162.3 · $10.00
Best value
- GPT-5.6 LunaECI 156.3 · $0.45
- Gemini 3.8 FlashECI 156.5 · $1.50
- GPT-5.6 TerraECI 159.1 · $4.50
ECI = Epoch Capabilities Index (Epoch AI). Price = blended USD per 1M tokens. Best value = no cheaper model scores higher.
Subscriptions and coding tools
All plans →AI subscriptions
- ChatGPTFree plan · Plus $20/mo
- ClaudeFree plan · Pro $20/mo
- Google AI (Gemini)Free plan · Google AI Pro $19.99/mo
- Mistral VibeFree plan · Pro $14.99/mo
AI coding tools
- GitHub CopilotFree plan · Pro $10/mo
- CursorFree plan · Pro $20/mo
- Claude CodeClaude Pro $20/mo
- OpenAI CodexFree plan · ChatGPT Plus $20/mo
- WindsurfFree plan · Pro $20/mo
- Google AntigravityFree plan · With Google AI Pro $19.99/mo
API pricing by provider
Subscription or API: which costs less? →
Is Claude Max or ChatGPT Pro worth it? See how many API messages each plan’s price buys.
Tokens to words converter →
How many words is 1,000 tokens? Convert either way, with pages and other languages.
Open-weight models, by provider →
What DeepInfra, Together, Fireworks and others charge to run gpt-oss, DeepSeek, Kimi, GLM, Qwen and Llama.
Image, video and voice APIs →
Veo 3.1, Gemini Omni, Nano Banana, GPT Image, Whisper and text-to-speech prices, per image, second or minute.
Every model, cheapest first
Model details →| Model | Provider | Input | Cached input | Output | Context |
|---|---|---|---|---|---|
| GPT-6 Luna | OpenAI | $0.10 | $0.01 | $0.50 | 1.05M |
| GPT-5.6 Luna | OpenAI | $0.20 | $0.02 | $1.20 | 1.05M |
| DeepSeek V4.1 Flash | DeepSeek | $0.30 | $0.006 | $1.20 | 1M |
| Gemini 3.5 Flash-Lite | $0.30 | $0.03 | $2.50 | 1.05M | |
| Gemini 2.5 Flash | $0.30 | $0.03 | $2.50 | 1.05M | |
| Gemini 3.8 Flash | $0.75 | $0.075 | $3.75 | 1.05M | |
| Grok 4.3 | xAI | $1.25 | $0.20 | $2.50 | 1M |
| DeepSeek V4 Pro | DeepSeek | $1.32 | $0.044 | $3.96 | 1M |
| Claude Haiku 4.5 | Anthropic | $1.00 | $0.10 | $5.00 | 200K |
| Grok 4.7 | xAI | $2.00 | $0.50 | $6.00 | 500K |
| Grok 4.6 | xAI | $2.00 | $0.50 | $6.00 | 500K |
| Gemini 3.5 Flash | $1.50 | $0.15 | $9.00 | 1.05M | |
| Gemini 2.5 Pro | $1.25 | $0.125 | $10.00 | 1.05M | |
| GPT-6 Sol | OpenAI | $2.00 | $0.20 | $10.00 | 1.05M |
| GPT-6.1 Sol | OpenAI | $2.00 | $0.10 | $10.00 | 1.05M |
| Claude Sonnet 5.5 | Anthropic | $2.00 | $0.20 | $10.00 | 1M |
| Claude Sonnet 5 | Anthropic | $2.00 | $0.20 | $10.00 | 1M |
| GPT-5.6 Terra | OpenAI | $2.00 | $0.20 | $12.00 | 1.05M |
| Gemini 3.1 ProPreview | $2.00 | $0.20 | $12.00 | 1.05M | |
| GPT-5.6 Sol | OpenAI | $4.00 | $0.40 | $20.00 | 1.05M |
| Claude Opus 5.5 | Anthropic | $4.00 | $0.20 | $20.00 | 1M |
| Claude Opus 5 | Anthropic | $5.00 | $0.50 | $25.00 | 1M |
| GPT-6 Astra | OpenAI | $10.00 | $1.00 | $50.00 | 1.05M |
| Claude Fable 5.1 | Anthropic | $10.00 | $0.25 | $50.00 | 1M |
USD per 1M tokens, standard tier. Sorted by blended price (3 input : 1 output).
Estimate your monthly API bill
- $55.00/mo$0.0011 / request
- GPT-5.6 LunaOpenAI$116.00/mo$0.0023 / request
- DeepSeek V4.1 FlashDeepSeek$156.00/mo$0.0031 / request
- Gemini 3.5 Flash-LiteGoogle$195.00/mo$0.0039 / request
- Gemini 2.5 FlashGoogle$195.00/mo$0.0039 / request
- Gemini 3.8 FlashGoogle$412.50/mo$0.0083 / request
- Claude Haiku 4.5Anthropic$550.00/mo$0.01 / request
- Grok 4.3xAI$575.00/mo$0.01 / request
- DeepSeek V4 ProDeepSeek$646.80/mo$0.01 / request
- Gemini 2.5 ProGoogle$800.00/mo$0.02 / request
- Gemini 3.5 FlashGoogle$870.00/mo$0.02 / request
- Grok 4.7xAI$980.00/mo$0.02 / request
- Grok 4.6xAI$980.00/mo$0.02 / request
- GPT-6 SolOpenAI$1,100/mo$0.02 / request
- GPT-6.1 SolOpenAI$1,100/mo$0.02 / request
- Claude Sonnet 5.5Anthropic$1,100/mo$0.02 / request
- Claude Sonnet 5Anthropic$1,100/mo$0.02 / request
- GPT-5.6 TerraOpenAI$1,160/mo$0.02 / request
- Gemini 3.1 ProGoogle$1,160/mo$0.02 / request
- GPT-5.6 SolOpenAI$2,200/mo$0.04 / request
- Claude Opus 5.5Anthropic$2,200/mo$0.04 / request
- Claude Opus 5Anthropic$2,750/mo$0.06 / request
- GPT-6 AstraOpenAI$5,500/mo$0.11 / request
- Claude Fable 5.1Anthropic$5,500/mo$0.11 / request
Standard-tier list prices. Long-context rates apply automatically where the provider publishes a threshold. Excludes cache-write surcharges, taxes, batch and volume discounts.
Pricing guides
All guides →GPT vs Claude API Cost: Every Current Model Compared
OpenAI’s GPT-6 and GPT-5.6 against Anthropic’s Claude Fable, Opus, Sonnet and Haiku — list prices, real-workload costs and the details that change the math.
The Cheapest AI APIs Right Now, Ranked
Every model from OpenAI, Anthropic, Google, xAI and DeepSeek ranked by blended price per token, plus the caveats that can flip the ranking.
How AI API Pricing Works: Tokens, Caching and Batch Explained
A plain-English guide to input and output tokens, prompt caching, batch discounts and long-context tiers — with a worked cost example.
How Many Tokens Do You Get with Claude Pro and Max?
Anthropic doesn’t publish a token allowance. How Claude’s session and weekly limits work, Pro vs Max 5x vs Max 20x, and what each plan’s price buys on the API.
How Much Does an AI Chatbot Cost for 10,000 Users?
A worked monthly bill for 200,000 chat replies on every tracked model, the cost per user, and four ways to cut it.
Best AI Model for Coding by Price: DeepSWE Scores vs Cost
Independent coding-benchmark scores next to the monthly cost of running a coding agent, plus when a flat-rate coding tool beats paying per token.
The Cheapest AI Model for RAG (Document Q&A)
RAG bills are driven by input and cached-input prices. Every model ranked for 50,000 document questions a month, plus the long-context threshold to watch.
Batch API Discounts Explained: 50% Off the Work That Can Wait
Which models offer a batch discount, what the discounted rate works out to, and which parts of a workload belong in the batch lane.