← News
API pricing

Cheapest AI APIs in 2026, ranked by output price

Leiolai 1 Research Fast is the cheapest AI API at $0.02 per million output tokens. Gemini 2.5 Flash-Lite is second at $0.40, and GPT-6 Luna is third at $0.50.

Three machined-metal podium blocks, with a silicon edge lighting the first-place block.
  1. 1Leiolai 1$0.02 output
  2. 2Gemini 2.5 Flash-Lite$0.40 output
  3. 3GPT-6 Luna$0.50 output
AI API prices per million output tokens.
  1. 1. Leiolai 1 Research Fast

    $0.01 input$0.02 output

    Leiolai 1 Research Fast costs $0.01 per million input tokens and $0.02 per million output tokens. It is the cheapest AI API.

  2. 2. Gemini 2.5 Flash-Lite

    $0.10 input$0.40 output

    Google charges $0.10 per million input tokens and $0.40 per million output tokens for Gemini 2.5 Flash-Lite. Gemini ranks second.

  3. 3. GPT-6 Luna

    $0.10 input$0.50 output

    OpenAI charges $0.10 per million input tokens and $0.50 per million output tokens for GPT-6 Luna. Luna ranks third.

  4. 4. Claude Haiku 4.5

    $1 input$5 output

    Anthropic charges $1 per million input tokens and $5 per million output tokens for Claude Haiku 4.5. Haiku ranks fourth.

  5. 5. Grok 4.7

    $2 input$6 output

    xAI charges $2 per million input tokens and $6 per million output tokens for Grok 4.7. Grok ranks fifth.

Full price comparison

Price per million tokensRanked by output
RankProvider and modelInputOutput
1Leiolai 1Leiolai · Research Fast$0.01$0.02
2Gemini 2.5 Flash-LiteGoogle$0.10$0.40
3GPT-6 LunaOpenAI$0.10$0.50
4Claude Haiku 4.5Anthropic$1$5
5Grok 4.7xAI$2$6

Prices were checked against each provider's official pricing on September 26, 2026.

Where GPT-6 Sol, Claude Opus, and Fable land

GPT-6 Sol, Claude Opus 5.5, and Claude Fable 5.1 cost more than the lowest-priced model from their respective providers. The GPT-6 Astra, Fable 5.1, and Leiolai 1 comparison covers context and output limits.

Call Leiolai 1 for $0.02 per million output tokens

Use leiolai-1 with reasoning_effort: "low" and mode: "non-private". The model stays the same; the Fast setting controls how much response depth to use.

EffortInput / 1MOutput / 1M
Low$0.01$0.02
Medium$0.25$0.65
High$6.40$21
Xhigh$160$700

Low, medium, and high Research rates are launch prices. Xhigh uses standard pricing. The Leiolai API pricing table carries the current rates and the API reference has a working Chat Completions request.

Check out the most expensive AI API rates in 2026.