AI Price Index: March 2026

Rebuilt on 2026-11-04 from the models.dev history · frozen, not revised afterwards · full index and methodology

In March 2026, the median AI model cost $1.225 per 1M tokens on labs' own APIs, 0% on the previous month. 13 models added, 0 retired, 0 repriced.

index101January 2026 = 100
median price$1.2250% vs last month
budget / premium$0.175 / $7.5010th / 90th percentile
tracked173models from 11 labs

// list price changes

No lab changed the list price of an existing model this month.

// added this month (13)

modellabinput $/1Moutput $/1Mblended
GPT-5.4 ProOpenAI$30.00$180.00$67.50
GPT-5.4OpenAI$2.50$15.00$5.625
Grok 4.20 (Non-Reasoning)xAI$2.00$6.00$3.00
Grok 4.20 (Reasoning)xAI$2.00$6.00$3.00
Grok 4.20 Multi-AgentxAI$2.00$6.00$3.00
GLM-5-TurboZ.AI (Zhipu)$1.20$4.00$1.90
GPT-5.4 miniOpenAI$0.75$4.50$1.6875
MiniMax-M2.7-highspeedMiniMax$0.60$2.40$1.05
Gemini 3.1 Flash Lite PreviewGoogle$0.25$1.50$0.5625
MiniMax-M2.7MiniMax$0.30$1.20$0.525
GPT-5.4 nanoOpenAI$0.20$1.25$0.4625
Mistral Small 4Mistral$0.15$0.60$0.2625
GLM-4.7-FlashXZ.AI (Zhipu)$0.07$0.40$0.1525

New in the catalog this month, which can include renamed models or older models added late.

// median price by lab, where it moved

labFebruary 2026March 2026change
xAI$2.725$3.00+10.1%
Mistral$0.75$0.725−3.3%
Google$0.85$0.7063−16.9%

// cheapest model by capability

capabilitymodelblended
tool calling, 128K+ contextCommand R7B$0.0656
reasoning and tool callingQwen Turbo$0.0875
1M+ contextGemini 1.5 Flash-8B$0.0656
image inputGemini 1.5 Flash-8B$0.0656

← February 2026April 2026 →full indexevery model (json)