What is LLM inference pricing?

LLM inference pricing refers to the cost of using Large Language Model APIs. Providers charge per token, with separate rates for input tokens (your prompts) and output tokens (model responses). Output tokens typically cost 2-5x more than input tokens.

How do I calculate LLM API costs?

LLM API costs are calculated by multiplying your token usage by the price per million tokens. For example, if you use 1M input tokens at $3/M and 500K output tokens at $15/M, your total cost would be $3 + $7.50 = $10.50.

Which LLM provider is cheapest?

The cheapest LLM provider depends on the model you need. For open-source models like Llama, providers like Together AI, Fireworks AI, and DeepInfra offer competitive rates. For proprietary models like GPT-4 or Claude, you'll typically need to use OpenAI or Anthropic directly.

What is a context window in LLMs?

A context window is the maximum number of tokens an LLM can process in a single request. Larger context windows (128K-1M+ tokens) allow for longer conversations and document analysis but may cost more. Most frontier models now support at least 128K tokens.

LLM Pricing

Compare per-token rates across 16 providers and 171 LLM models

Updated daily

💡Top Picks Right Now

Click to filter table

Price Trends

Moving average across providers (per 1M tokens) Input Output

Cheapest Option

Qwen 3.5 27B

$0.216 / $1.12per 1M tokens

Providers: 3

Most Popular

GPT-OSS-120B

$0.124 / $0.503per 1M tokens

Providers: 7

Top Performance

Claude Opus 4.5

$3.81 / $19.05per 1M tokens

Providers: 2

Pricing

Showing 15 of 168 models (403 price records)

Model ↕	Creator ↕	Modalities	Avg Input/1M ↕	Input Range ↕	Speed ↕	$/Perf ↕	Providers ↓
GPT-OSS-120B 128K ctxJan 2025	OpenAI	Aa	$0.119/1M	$0.039 – $0.176	218 t/s	$0.0009	10
Llama 3.3 70B 128K ctxMar 2024	Meta	Aa	$0.606/1M	$0.100 – $1.05	-	-	7
Mistral Small 32K ctxSep 2023	Mistral	Aa	$0.272/1M	$0.050 – $1.00	98.6 t/s	$0.0008	7
Llama 3.1 8B 128K ctxDec 2023	Meta	Aa	$0.135/1M	$0.020 – $0.235	0.0 t/s	-	7
DeepSeek R1 64K ctxNov 2024	DeepSeek	Aa	$1.66/1M	$0.320 – $3.75	0.0 t/s	-	6
DeepSeek V3 64K ctxJun 2024	DeepSeek	Aa	$0.572/1M	$0.200 – $1.25	0.0 t/s	-	6
GPT-OSS-20B 128K ctxJan 2025	OpenAI	Aa	$0.054/1M	$0.030 – $0.075	317 t/s	$0.0004	6
Qwen 3 235B 128K ctx	Alibaba	Aa	$0.456/1M	$0.071 – $0.878	67.0 t/s	$0.0015	5
DeepSeek V3.1 33K ctx	DeepSeek	Aa	$0.423/1M	$0.150 – $0.600	0.0 t/s	-	5
GLM-5 128K ctx	Zhipu	Aa	$0.840/1M	$0.600 – $1.00	0.0 t/s	-	5
Qwen 3 Coder 30B 128K ctx	Alibaba	Aa	$0.535/1M	$0.070 – $2.00	112 t/s	$0.0024	5
Kimi K2.5 128K ctx	Moonshot	Aa	$0.518/1M	$0.440 – $0.600	-	-	5
GLM-4.7 128K ctx	Zhipu	Aa	$0.380/1M	$0.070 – $0.600	129 t/s	$0.0031	5
Llama 4 Scout 328K ctx	Meta	Aa	$0.124/1M	$0.080 – $0.180	136 t/s	$0.0022	5
Kimi K2 128K ctx	Moonshot	Aa	$0.814/1M	$0.570 – $1.20	32.1 t/s	$0.072	5

Understanding LLM Pricing

Input vs Output Tokens

LLM APIs charge separately for input tokens (your prompts) and output tokens (model responses). Output tokens are usually 2-5x more expensive than input tokens.

Context Windows

The context window determines how much text a model can process at once. Larger context windows allow for longer conversations and document analysis.

Open vs Proprietary Models

Open-source models (Llama, Mistral) are often cheaper but may require more tuning. Proprietary models (GPT-4, Claude) typically offer better out-of-box performance.

Batch Discounts

Many providers offer 50% discounts for batch/async API usage. Consider batch APIs for non-time-sensitive workloads to reduce costs.