Pay as you go
Pricing
The API is pay as you go: top up your balance and pay per token at each model's price. flua Chat is free within your plan's limits, and a monthly plan opens strong models in the chat and discounts the API.
- 01
Pay as you go
You are only charged for the tokens the model actually processed.
- 02
One balance
Your balance works for every model in the catalog.
- 03
Prices upfront
Every model's price is published here and in the catalog.
How a request is priced
(input tokens × input price + output tokens × output price) ÷ 1,000,000
Example
Claude Opus 5.5: 1,500 input and 500 output tokens
(1,500 × $2.00168 ₽90,00 ₴ + 500 × $8.00672 ₽360 ₴) ÷ 1M = $0.0070,59 ₽0,32 ₴
Cost calculator
- One request
- $0.007
- Per day
- $1.40
- Per month
- $42.00
Subscriptions
Chat models on a plan, the API at a discount
A plan opens models in flua Chat with no per-request charge — with 5-hour and weekly limits — and discounts the API. Paid from your balance, cancel any time.
Free
Fast models in the chat for free
Free
- List prices
- 60 requests per minute per key
- up to 5 requests at once
- up to 3 API keys
Free in the chat
Each model has its own 5-hour and weekly limit.
- Qwen3.8 Flash≈120/wk
- GLM-5.3 Flash≈80/wk
- HY 4 Preview≈60/wk
- DeepSeek V4.1 Flash≈60/wk
- DeepSeek V4.1 Flash 1017≈60/wk
8 GPT Image 2.5 pictures a day
Claude Opus 5.5 via the API on this plan:
$2.00168 ₽90,00 ₴ / $8.00672 ₽360 ₴ · 1M
Plus
Claude Haiku and GPT-6 Luna in the chat
$5.00420 ₽225 ₴/ mo
- −10% on every model
- 180 requests per minute per key
- up to 10 requests at once
- up to 10 API keys
Free in the chat
One limit for all models: the figure is messages a week if you use only that model.
- Claude Haiku 5.5up to ≈2,700/wk
- GPT-6 Lunaup to ≈2,600/wk
- Qwen3.8 Flashup to ≈2,100/wk
- GLM-5.3 Flashup to ≈1,200/wk
- HY 4 Previewup to ≈1,000/wk
and 2 more
- DeepSeek V4.1 Flashup to ≈910/wk
- DeepSeek V4.1 Flash 1017up to ≈43/wk
25 GPT Image 2.5 pictures a day
Claude Opus 5.5 via the API on this plan:
$1.80151 ₽81,00 ₴ / $7.20605 ₽324 ₴ · 1M
Pro
Best valueOpus, Sonnet, GPT-6 and Gemini in the chat
$15.001 260 ₽675 ₴/ mo
- −20% on every model
- 600 requests per minute per key
- up to 20 requests at once
- up to 30 API keys
Free in the chat
One limit for all models: the figure is messages a week if you use only that model.
- Claude Opus 5.5up to ≈150/wk
- Claude Sonnet 5.5up to ≈250/wk
- GPT-6 Astraup to ≈270/wk
- Gemini 4 Argonup to ≈460/wk
- Gemini 3.1 Proup to ≈920/wk
and 10 more
- GPT-6.1 Solup to ≈3,700/wk
- Grok 4.7up to ≈7,800/wk
- Gemini 3.8 Flashup to ≈2,700/wk
- Claude Haiku 5.5up to ≈7,700/wk
- GPT-6 Lunaup to ≈7,500/wk
- Qwen3.8 Flashup to ≈6,000/wk
- GLM-5.3 Flashup to ≈3,400/wk
- HY 4 Previewup to ≈2,900/wk
- DeepSeek V4.1 Flashup to ≈2,600/wk
- DeepSeek V4.1 Flash 1017up to ≈120/wk
50 GPT Image 2.5 pictures a day
up to 4 pictures at once
Claude Opus 5.5 via the API on this plan:
$1.60134 ₽72,00 ₴ / $6.40538 ₽288 ₴ · 1M
Max
Three times the limits on everything
$40.003 360 ₽1 800 ₴/ mo
- −30% on every model
- 2000 requests per minute per key
- up to 50 requests at once
- up to 100 API keys
Free in the chat
One limit for all models: the figure is messages a week if you use only that model.
- Claude Opus 5.5up to ≈430/wk
- Claude Sonnet 5.5up to ≈720/wk
- GPT-6 Astraup to ≈760/wk
- Gemini 4 Argonup to ≈1,200/wk
- Gemini 3.1 Proup to ≈2,500/wk
and 10 more
- GPT-6.1 Solup to ≈10,000/wk
- Grok 4.7up to ≈21,000/wk
- Gemini 3.8 Flashup to ≈7,700/wk
- Claude Haiku 5.5up to ≈21,000/wk
- GPT-6 Lunaup to ≈21,000/wk
- Qwen3.8 Flashup to ≈16,000/wk
- GLM-5.3 Flashup to ≈9,600/wk
- HY 4 Previewup to ≈8,300/wk
- DeepSeek V4.1 Flashup to ≈7,200/wk
- DeepSeek V4.1 Flash 1017up to ≈350/wk
100 GPT Image 2.5 pictures a day
up to 4 pictures at once
Claude Opus 5.5 via the API on this plan:
$1.40118 ₽63,00 ₴ / $5.60470 ₽252 ₴ · 1M
The discount applies to every model, but a price never drops below cost — a few cheap models may get less than the full discount.
All prices
- Text$0.054,20 ₽2,25 ₴$0.0050,42 ₽0,23 ₴$0.2521,00 ₽11,25 ₴Claude Haiku 5.5−50%claude-haiku-5-5
- Text$2.00168 ₽90,00 ₴$0.2016,80 ₽9,00 ₴$8.00672 ₽360 ₴Claude Opus 5.5−50%claude-opus-5-5
- Text$1.20101 ₽54,00 ₴$0.1210,08 ₽5,40 ₴$4.00336 ₽180 ₴claude-sonnet-5-5
- Text$4.00336 ₽180 ₴$0.4033,60 ₽18,00 ₴$16.001 344 ₽720 ₴Claude Fable 5.1−60%claude-fable-5-1
- Text$1.50126 ₽67,50 ₴$0.1512,60 ₽6,75 ₴$7.50630 ₽338 ₴claude-sonnet-4-6
- Text$0.7563,00 ₽33,75 ₴$0.0756,30 ₽3,38 ₴$4.00336 ₽180 ₴GPT-6.1 Sol−60%gpt-6.1-sol
- Text$2.00168 ₽90,00 ₴$0.2016,80 ₽9,00 ₴$10.00840 ₽450 ₴GPT-6 Astra−80%gpt-6-astra
- Text$0.054,20 ₽2,25 ₴$0.0050,42 ₽0,23 ₴$0.2016,80 ₽9,00 ₴GPT-6 Luna−50%gpt-6-luna
- Images$0.010,84 ₽0,45 ₴/ requestgpt-image-2.5-flare
- Images$0.010,84 ₽0,45 ₴/ requestgpt-image-2.5-sunburst
- Embeddings$0.1512,60 ₽6,75 ₴——text-embedding-3-large
- Embeddings$0.0161,34 ₽0,72 ₴——text-embedding-3-small
- Embeddings$0.0957,98 ₽4,28 ₴——text-embedding-ada-002
- Text$1.0084,00 ₽45,00 ₴$0.108,40 ₽4,50 ₴$2.50210 ₽113 ₴Gemini 4 ArgonFastgemini-4-argon
- Text$0.3529,40 ₽15,75 ₴$0.0352,94 ₽1,57 ₴$1.20101 ₽54,00 ₴gemini-3.8-flash
- Text$1.0084,00 ₽45,00 ₴$0.108,40 ₽4,50 ₴$3.00252 ₽135 ₴gemini-3.1-pro-preview
- Images$0.075,88 ₽3,15 ₴/ requestnano-banana-2.1
- Images$0.075,88 ₽3,15 ₴/ requestnano-banana-2
- Images$0.054,20 ₽2,25 ₴/ requestnano-banana-pro
- Embeddings$0.2218,48 ₽9,90 ₴——gemini-embedding-2-preview
- Text$0.7563,00 ₽33,75 ₴$0.1915,75 ₽8,44 ₴$2.00168 ₽90,00 ₴Grok 4.7−63%grok-4.7
- Images$0.043,36 ₽1,80 ₴/ requestgrok-imagine-image
- Text$0.108,40 ₽4,50 ₴$0.010,84 ₽0,45 ₴$0.3025,20 ₽13,50 ₴deepseek-v4.1-flash
- Text$0.108,40 ₽4,50 ₴$0.010,84 ₽0,45 ₴$0.3025,20 ₽13,50 ₴deepseek-v4.1-flash-1017
- Text$0.0756,30 ₽3,38 ₴$0.0151,26 ₽0,67 ₴$0.1512,60 ₽6,75 ₴qwen3.8-flash
- Text$0.5042,00 ₽22,50 ₴$0.108,40 ₽4,50 ₴$2.00168 ₽90,00 ₴Qwen3.8 Max−67%qwen3.8-max
- Text$0.0756,30 ₽3,38 ₴$0.0151,26 ₽0,67 ₴$0.1512,60 ₽6,75 ₴qwen3.8-omni-flash
- Text$1.50126 ₽67,50 ₴$0.3831,50 ₽16,88 ₴$7.00588 ₽315 ₴kimi-k3.1-pro
- Text$1.50126 ₽67,50 ₴$0.3831,50 ₽16,88 ₴$7.00588 ₽315 ₴Kimi K3−50%kimi-k3
- Text$0.7058,80 ₽31,50 ₴$0.1411,76 ₽6,30 ₴$2.20185 ₽99,00 ₴GLM-5.3−50%glm-5.3
- Text$0.108,40 ₽4,50 ₴$0.021,68 ₽0,90 ₴$0.2521,00 ₽11,25 ₴glm-5.3-flash
- Text$0.1512,60 ₽6,75 ₴$0.0151,26 ₽0,67 ₴$0.6050,40 ₽27,00 ₴minimax-m3
- Text$0.108,40 ₽4,50 ₴$0.010,84 ₽0,45 ₴$0.2016,80 ₽9,00 ₴mimo-v2.6-flash
- Text$0.3025,20 ₽13,50 ₴$0.032,52 ₽1,35 ₴$0.6050,40 ₽27,00 ₴MiMo V2.6 Pro−31%mimo-v2.6-pro
- Text$0.2016,80 ₽9,00 ₴$0.054,20 ₽2,25 ₴$1.20101 ₽54,00 ₴hy4-preview
- Text$1.50126 ₽67,50 ₴$0.3831,50 ₽16,88 ₴$5.00420 ₽225 ₴Muse Spark 1.3Fastmuse-spark-1.3
- Text$1.20101 ₽54,00 ₴$0.3025,20 ₽13,50 ₴$5.00420 ₽225 ₴muse-wonder
Showing 37 of 37
About billing
What is a token?
A chunk of text the model works with. For English, 1,000 tokens is roughly 750 words; other languages usually take more tokens.
How do I see what a request cost?
Every API response has a usage field with token counts, and the dashboard keeps a request log with the cost of each call.
Why do some models have two prices?
Input tokens (your prompt and the conversation history) and output tokens (the model's answer) are priced differently. Some models also apply a multiplier to very long prompts — it is listed on the model page.
Is there prompt caching?
Yes. When the start of a request repeats — a system prompt, documents, the conversation so far — the model reads it from cache, and those input tokens cost what the model's maker charges: 10% of the normal input price (ten times less) for Claude, GPT, Gemini, DeepSeek, MiniMax and MiMo, 20% for Qwen and GLM, 25% for Grok, Kimi and Hunyuan. OpenAI, Google, DeepSeek and other models cache on their own; for Claude flua turns caching on automatically, and writing to Claude's cache costs 125% of the input price. The usage of every response and the request log show how many tokens came from cache.
What does a plan give in the chat?
Every plan, the free one included, lets you use some models in flua Chat with no per-request charge — within 5-hour and weekly limits. Free has fast models (Qwen, GLM, DeepSeek, HY) and 8 pictures a day; Plus adds Claude Haiku and GPT-6 Luna; Pro and Max add Claude Opus and Sonnet, GPT-6, Gemini and Grok, up to 4 pictures at once. When a limit runs out, switch the chat to your balance and go on at API prices. Requests through API keys have no limits and are paid from the balance with the plan's discount.
Do prices change?
Prices may follow the providers' own price changes. The current price is always on this page and in the catalog.