Skip to content
AZRouter.dev

Pricing

Every model’s price, next to OpenRouter’s

AZRouter bills per unit in US dollars from prepaid credits; a model OpenRouter does not carry is compared with its provider's list price.
Save up to 10% against the reference price

Price per million tokens

12 models billed per token, 1 per character, 9 providers, OpenRouter prices as of

  • Text Embedding 3 Largeopenai/text-embedding-3-large
    Input
    Input $0.143
    Output
    Output: not billed
    Cache read
    Cache read: not billed

    Blended$0.143input only

    OpenRouter$0.13+10%

  • Claude Haiku 5.5Tiered above 100Kanthropic/claude-haiku-5.5
    Input
    Input $0.11
    Output
    Output $0.55
    Cache read
    Cache read $0.011

    Blended$0.22

    OpenRouter$0.20+10%

  • GPT-6 LunaTiered above 272Kopenai/gpt-6-luna
    Input
    Input $0.11
    Output
    Output $0.55
    Cache read
    Cache read $0.011

    Blended$0.22

    OpenRouter$0.20+10%

  • Qwen3.8 Flashqwen/qwen3.8-flash
    Input
    Input $0.165
    Output
    Output $0.517
    Cache read
    Cache read $0.0176

    Blended$0.253

    OpenRouter$0.23+10%

  • Mistral Large 4mistralai/mistral-large-4-0
    Input
    Input $0.748
    Output
    Output $2.30
    Cache read
    Cache read $0.077

    Blended$1.14

    OpenRouter$1.03+10%

  • GLM 5.3z-ai/glm-5.3
    Input
    Input $0.0869
    Output
    Output $5.28
    Cache read
    Cache read $0.0858

    Blended$1.39

    OpenRouter$1.26+10%

  • Gemini 3.8 Flashgoogle/gemini-3.8-flash
    Input
    Input $0.7425
    Output
    Output $3.71
    Cache read
    Cache read $0.07425

    Blended$1.48Save 10%

    OpenRouter$1.65−10%

  • Grok 4.7Tiered from 200Kx-ai/grok-4.7
    Input
    Input $2.20
    Output
    Output $6.60
    Cache read
    Cache read $0.55

    Blended$3.30

    OpenRouter$3.00+10%

  • Kimi K3moonshotai/kimi-k3
    Input
    Input $0.3344
    Output
    Output $13.79
    Cache read
    Cache read $0.231

    Blended$3.70

    OpenRouter$3.36+10%

  • Gemini 4 Argongoogle/gemini-4-argon
    Input
    Input $2.20
    Output
    Output $11.00
    Cache read
    Cache read $0.11

    Blended$4.40

    OpenRouter$4.00+10%

  • GPT-6.1 SolTiered above 272Kopenai/gpt-6.1-sol
    Input
    Input $2.20
    Output
    Output $11.00
    Cache read
    Cache read $0.11

    Blended$4.40

    OpenRouter$4.00+10%

  • Claude Opus 5.5anthropic/claude-opus-5.5
    Input
    Input $4.40
    Output
    Output $22.00
    Cache read
    Cache read $0.22

    Blended$8.80

    OpenRouter$8.00+10%

One blended million
750K input250K output

Blended = (3 × input + output) ÷ 4, the cost of one million tokens at three input tokens per output token, so models compare on one number. You are billed per unit used, not by the blend. An embedding model bills input tokens only, so its blended price is its input price.

Tiered models are shown at their base rate. A prompt past the threshold bills the whole request at the long-context rate; the model page lists both.

Each request's total is rounded to the nearest $0.000002 (halves up), so it can differ from the per-unit math by up to $0.000001. How rounding works

What a month could cost

An estimate at list prices: the blended price × tokens per month, at three input tokens per output token. The 6 lowest blended prices, cheapest first; the filter above applies here too.

  • Text Embedding 3 Largeinput only+10%
    1M / mo
    $0.143vsOpenRouter $0.13
    10M / mo
    $1.43vsOpenRouter $1.30
    100M / mo
    $14.30vsOpenRouter $13.00
  • Claude Haiku 5.5Tiered above 100K+10%
    1M / mo
    $0.22vsOpenRouter $0.20
    10M / mo
    $2.20vsOpenRouter $2.00
    100M / mo
    $22.00vsOpenRouter $20.00
  • GPT-6 LunaTiered above 272K+10%
    1M / mo
    $0.22vsOpenRouter $0.20
    10M / mo
    $2.20vsOpenRouter $2.00
    100M / mo
    $22.00vsOpenRouter $20.00
  • Qwen3.8 Flash+10%
    1M / mo
    $0.253vsOpenRouter $0.23
    10M / mo
    $2.53vsOpenRouter $2.30
    100M / mo
    $25.30vsOpenRouter $23.00
  • Mistral Large 4+10%
    1M / mo
    $1.14vsOpenRouter $1.03
    10M / mo
    $11.36vsOpenRouter $10.33
    100M / mo
    $113.57vsOpenRouter $103.25
  • GLM 5.3+10%
    1M / mo
    $1.39vsOpenRouter $1.26
    10M / mo
    $13.85vsOpenRouter $12.59
    100M / mo
    $138.52vsOpenRouter $125.93

Estimates, not quotes: your bill depends on your real input and output mix, cached tokens and the per-request rounding. Tiered models are estimated at their base rate; a prompt past the threshold bills the whole request at the long-context rate.

Other units

Models billed per character rather than per token, each against its own reference. No blended price: these units do not mix with tokens.

  • Eleven v3elevenlabs/eleven-v3

    1M characters$44.00

    OpenRouter$40.00+10%

Pricing questions

Short answers. The docs have the details, with worked examples.

Read the docs
What are the prices compared with?
Every price is per unit, in US dollars, next to OpenRouter's public price for the same model and unit, with the difference in percent. A model OpenRouter does not carry is compared with its provider's own list price instead.
Is every model marked up the same?
No. Each model has its own markup over its reference price, so the difference is shown model by model, up or down.
How do I pay?
AZRouter is prepaid: you top up credits in US dollars, and each call debits its cost from your balance.
How is a request rounded?
Each request's total is rounded to the nearest $0.000002, halves up, so it can differ from the per-unit math by up to $0.000001.
How current are the OpenRouter prices?
OpenRouter's prices were last checked . A model compared with its provider's list price says so on its model page.