Skip to content
FullsendAI
Log inGet started

Price comparison

LLM API price comparison

What the same model costs from seven providers, and the fees each adds. FullsendAI's prices are read live from our catalogue; everyone else's carry the date we checked them (Nov 4, 2026 or earlier), and a price older than 30 days is replaced by a link to their pricing.

Get your API keySee our pricingLast verified Oct 15, 2026

Prices per million tokens

Input / output, USD. For OpenRouter: the price most of its providers charge, with the lowest-priced endpoint (and its precision) beneath. For Requesty: the route named. “Not listed” means the model wasn't on that provider's list on the date shown.

ModelFullsendAIOpenRouterRequestyDeepInfraTogether AIFireworks AIGroq
DeepSeek V4.1 FlashDeepSeekNot offered yet$0.30 / $1.20from $0.022 / $0.192 (fp8, Morph)Source · $0.30 / $1.20Source · $0.14 / $0.42Promotional priceSource · $0.30 / $1.20Source · $0.30 / $1.20Source · Not listedSource ·
GLM 5.3 FlashZ.aiNot offered yet$0.15 / $0.50from $0.0189 / $0.0625 (fp4, OpenInference)Source · $0.15 / $0.50Source · $0.075 / $0.25Promotional priceSource · $0.15 / $0.50Source · $0.15 / $0.50Source · Not listedSource ·
GPT OSS 120BOpenAINot offered yet$0.15 / $0.60from $0.03 / $0.15 (fp4, Near AI)Source · $0.15 / $0.60Source · $0.037 / $0.17Source · $0.15 / $0.60Source · $0.15 / $0.60Source · $0.15 / $0.60Source ·
GPT OSS 20BOpenAINot offered yet$0.07 / $0.15from $0.018 / $0.09 (fp8, Darkbloom)Source · $0.07 / $0.30Source · $0.03 / $0.14Source · Not listedSource · Not listedSource · $0.075 / $0.30Source ·
Kimi K2.6Moonshot AINot offered yet$0.95 / $4.00from $0.39 / $1.80 (int4, Chutes)Source · $0.95 / $4.00Source · $0.75 / $3.50Source · Not listedSource · Not listedSource · Not listedSource ·
Qwen3.8 MaxAlibabaNot offered yet$2.00 / $6.00Source · $2.00 / $6.00Source · $1.65 / $4.951Source · Not listedSource · $2.00 / $6.00Source · Not listedSource ·
DeepSeek V4 Pro 0813DeepSeekNot offered yet$1.32 / $3.96from $0.675 / $1.20 (fp4, Sail Research)Source · $1.32 / $3.96Source · $1.30 / $2.60Source · $1.32 / $3.96Source · Not listedSource · Not listedSource ·
Claude Opus 5.5AnthropicNot offered yet$4.00 / $20.00Source · $4.00 / $20.00Source · $4.00 / $20.00Source · Not listedSource · Not listedSource · Not listedSource ·
Gemini 3.1 ProGoogleNot offered yet$2.00 / $12.00Source · $2.00 / $12.00Source · $2.00 / $12.00Source · Not listedSource · Not listedSource · Not listedSource ·

Fees on top of the price

What each charges beyond the per-token price, from its own pricing page or FAQ.

  • FullsendAI

    On usage
    None
    Card top-up
    None
    Crypto top-up
    None

  • OpenRouter

    On usage
    None stated
    Card top-up
    5.5%, $0.80 minimum
    Crypto top-up
    5%
    Source
    Source ·

  • Requesty

    On usage
    5% on model cost
    Card top-up
    None stated
    Crypto top-up
    No crypto option listed
    Source
    Source ·

  • DeepInfra

    On usage
    None stated
    Card top-up
    None stated
    Crypto top-up
    No crypto option listed
    Source
    Source ·

  • Together AI

    On usage
    None stated
    Card top-up
    None stated
    Crypto top-up
    No crypto option listed
    Source
    Source ·

  • Fireworks AI

    On usage
    None stated
    Card top-up
    None stated
    Crypto top-up
    No crypto option listed
    Source
    Source ·

  • Groq

    On usage
    None stated
    Card top-up
    None stated
    Crypto top-up
    No crypto option listed
    Source
    Source ·
ProviderOn usageCard top-upCrypto top-upSource
FullsendAINoneNoneNoneOur pricing
OpenRouterNone stated5.5%, $0.80 minimum5%Source ·

Fees are charged when you buy credits, not per request.

Requesty5% on model costNone statedNo crypto option listedSource ·

“You pay for what your application spends, plus 5%.”

DeepInfraNone statedNone statedNo crypto option listedSource ·

No platform or top-up fee stated.

Together AINone statedNone statedNo crypto option listedSource ·

No platform or top-up fee stated.

Fireworks AINone statedNone statedNo crypto option listedSource ·

No platform or top-up fee stated.

GroqNone statedNone statedNo crypto option listedSource ·

No platform or top-up fee stated.

What a month costs, fees included

100M input and 10M output tokens, no cache hits, one card top-up for exactly the cost. Only models every provider below lists are shown.

No model has a current price on every side today, so there is no like-for-like example to show. See the price table above.

Reading the table

  • Same model, same version: we never compare a model with a “close enough” one.
  • Promotions count: a provider's discounted price is shown while it runs, with a note.
  • Precision matters: routers' lowest prices are often quantized (fp4, int4). We show the precision when the provider publishes it.
  • Cache and batch discounts aren't in these numbers; coding agents with many cache hits pay less everywhere that discounts cached input.

Questions

Which LLM API is cheapest?

It depends on the model and on fees, and it changes weekly. Read the table for the model you use; each price links to its source with the date we checked it. We don't claim FullsendAI is the cheapest.

How often is this updated?

FullsendAI's prices are live. Others are refreshed weekly from the providers' public APIs where they publish one, and checked by hand otherwise. See the methodology.

Why do router prices show two numbers?

A router like OpenRouter sends requests for one model to many providers, at different prices and precisions. We show the price most of them charge and the lowest-priced endpoint.

Related comparisons

One key. Every model. No top-up fee.

Works with the OpenAI and Anthropic SDKs, Claude Code, Codex and other coding agents. Every model runs in the USA.

Get your API keyRead the quickstart

OpenRouter, Requesty, DeepInfra, Together AI, Fireworks AI, Groq, OpenAI, Anthropic, Claude, Google, Gemini and DeepSeek are trademarks of their respective owners. FullsendAI is not affiliated with, endorsed by or sponsored by any of them; names are used only to identify the products compared.

Based on publicly available information as of the dates shown; features and prices change. How we compare · Report an inaccuracy