AI models and their prices,
compared for you.
300+ text models, 212 providers and 6600+ prices, read every day from public catalogs. See what a model costs from the lab that made it, who sells it for less, and the reason behind every price below list.
Cheapest today for DeepSeek V4 Flash 0731
OpenRouter$0.04 / $0.10 per M tokens in / out
$0.22/M tokens
cheapest price we can explain
$1.05DeepSeek list price
Routes to a host serving the open weights; precision not disclosed.
prices as of · re-ranked daily
Top 10 cheapest models
full ranking →32k context or more, sold by 3 providers or more · prices as of
- 1Llama 3.2 1B Instruct
Inference · 60k context · 11 providers · $0.01 / $0.01 in / out$0.04/M tokens - 2Llama 3.2 3B InstructInference · 131k context · 17 providers · $0.02 / $0.02 in / out$0.08/M tokens
- 3Mistral Nemo
Nous Portal · 128k context · 25 providers · $0.02 / $0.03 in / out$0.08/M tokens - 4GLM-5.3-Flash
Merge Gateway · 1M context · 76 providers · $0.01 / $0.05 in / out$0.10/M tokens - 5Llama 3.1 8B Instruct
Kilo Gateway · 131k context · 24 providers · $0.02 / $0.04 in / out$0.10/M tokens - 6Mercury 2.5
NanoGPT · 260k context · 8 providers · $0.04 / $0.004 in / out$0.12/M tokens - 7Ling 3.0 FlashOpenRouter · 262k context · 12 providers · $0.02 / $0.06 in / out$0.13/M tokens
- 8Qwen3.7 Flash
Eden AI · 1M context · 15 providers · $0.02 / $0.08 in / out$0.14/M tokens - 9gpt-oss-20bKilo Gateway · 131k context · 43 providers · $0.02 / $0.10 in / out$0.16/M tokens
- 10Ministral 3B (latest)
Azure Cognitive Services · 128k context · 14 providers · $0.04 / $0.04 in / out$0.16/M tokens
Top 10 biggest savings against list
full ranking →below the list price, with a published reason · prices as of
- 1Nano BananaKilo Gateway · Google lists it at $30.90$1.70/M tokens94% off list
- 2Qwen-VL OCREden AI · Alibaba Cloud lists it at $2.88$0.24/M tokens92% off list
- 3Qwen-MT Plus
Jiekou · Alibaba Cloud lists it at $14.75$1.50/M tokens90% off list - 4Qwen3 32BOpenRouter · Alibaba Cloud lists it at $4.90$0.52/M tokens89% off list
- 5Qwen2.5 72B Instruct
Requesty · Alibaba Cloud lists it at $9.80$1.09/M tokens89% off list - 6Qwen3 235B-A22BOpenRouter · Alibaba Cloud lists it at $4.90$0.61/M tokens88% off list
- 7Qwen3.8 27BNanoGPT · Alibaba Cloud lists it at $3.82$0.49/M tokens87% off list
- 8Qwen3-Coder 30B-A3B Instruct
Cortecs · Alibaba Cloud lists it at $3.60$0.46/M tokens87% off list - 9Mistral NemoNous Portal · Mistral AI lists it at $0.60$0.08/M tokens86% off list
- 10Qwen3-Next 80B-A3B (Thinking)Eden AI · Alibaba Cloud lists it at $7.50$1.07/M tokens86% off list
Cheapest with tool calling
Mistral Nemo
Nous Portal$0.08/M tokens
128k context · Nous Portal
can call your own functions
Cheapest with 1M context
GLM-5.3-Flash
Merge Gateway$0.10/M tokens
1M context · Merge Gateway
a million tokens in one call
Index
337
text models · 161 with open weights
212
providers
6617
prices
24h
between reads
What the makers charge for a million output tokens (USD)
This week
Every price that
moved, dated
what changed →
Providers with the most models
What are you paying a model to do?
Today's rankings, recomputed daily
Every ranking uses the standard price: three parts input to one part output per million tokens, at each provider's standard rate, with no cache discount and no batch rate. A price below the maker's own with nothing published to account for it is shown on the model's page and ranked nowhere.