AI Pricing Changelog

Track how AI API pricing changes over time. We snapshot pricing data daily and record every change.

How it works: Every day at 06:00 UTC we snapshot pricing data from all 103 models across 11 providers. When a price changes, it appears here automatically. Data available via the price history API.

(4 changes)

Model Field Old New Change
Gemini 2.5 Flash-Lite Model -- added --
Gemini 2.0 Flash Model -- added --
Gemini 2.0 Flash-Lite Model -- added --
Gemini 3.5 Flash Model -- added --

(9 changes)

Model Field Old New Change
Ministral 3B Input / 1M $0.0400 $0.1000 +150.0%
Ministral 3B Output / 1M $0.0400 $0.1000 +150.0%
Ministral 8B Input / 1M $0.1000 $0.1500 +50.0%
Ministral 8B Output / 1M $0.1000 $0.1500 +50.0%
Mistral Small 4 Input / 1M $0.1500 $0.1000 -33.3%
Mistral Small 4 Output / 1M $0.6000 $0.3000 -50.0%
Qwen3.7-Max (Together) Input / 1M $2.5000 $1.2500 -50.0%
Qwen3.7-Max (Together) Cached Input / 1M -- $0.1300 --
Qwen3.7-Max (Together) Output / 1M $7.5000 $3.7500 -50.0%

(1 change)

Model Field Old New Change
Claude Opus 4.8 Model -- added --

(1 change)

Model Field Old New Change
Qwen3.7-Max (Together) Model -- added --

(1 change)

Model Field Old New Change
Llama 4 Maverick Input / 1M $0.2000 $0.1500 -25.0%

(8 changes)

Model Field Old New Change
GPT-5.2 Model -- added --
GPT-5.2 Pro Model -- added --
GPT-5.1 Model -- added --
GPT-5 Model -- added --
GPT-5 Pro Model -- added --
GPT-5 mini Model -- added --
GPT-5 nano Model -- added --
o3-pro Model -- added --

(9 changes)

Model Field Old New Change
Openai/gpt Oss 120b Output / 1M $0.7500 $0.6000 -20.0%
Openai/gpt Oss 20b Input / 1M $0.1000 $0.0750 -25.0%
Openai/gpt Oss 20b Cached Input / 1M $0.0500 $0.0375 -25.0%
Openai/gpt Oss 20b Output / 1M $0.5000 $0.3000 -40.0%
Grok 4.20 Input / 1M $2.0000 $1.2500 -37.5%
Grok 4.20 Output / 1M $6.0000 $2.5000 -58.3%
GPT OSS Safeguard 20B Model -- added --
Llama 4 Scout 17B 16E Instruct Model -- added --
Qwen3 32B Model -- added --

(12 changes)

Model Field Old New Change
Openai/gpt Oss 120b Cached Input / 1M -- $0.0750 --
Openai/gpt Oss 120b Output / 1M $0.6000 $0.7500 +25.0%
Openai/gpt Oss 20b Input / 1M $0.0750 $0.1000 +33.3%
Openai/gpt Oss 20b Cached Input / 1M -- $0.0500 --
Openai/gpt Oss 20b Output / 1M $0.3000 $0.5000 +66.7%
Llama 4 Maverick Output / 1M $0.2000 $0.6000 +200.0%
Llama 4 Scout Input / 1M $0.1500 $0.0800 -46.7%
Llama 4 Scout Output / 1M $0.1500 $0.3000 +100.0%
Ministral 3B Input / 1M $0.1000 $0.0400 -60.0%
Ministral 3B Output / 1M $0.1000 $0.0400 -60.0%
Ministral 8B Input / 1M $0.1500 $0.1000 -33.3%
Ministral 8B Output / 1M $0.1500 $0.1000 -33.3%

(22 changes)

Model Field Old New Change
Ministral 3B Input / 1M $0.0400 $0.1000 +150.0%
Ministral 3B Output / 1M $0.0400 $0.1000 +150.0%
Ministral 8B Input / 1M $0.1000 $0.1500 +50.0%
Ministral 8B Output / 1M $0.1000 $0.1500 +50.0%
Mistral Small 4 Input / 1M $0.1000 $0.1500 +50.0%
Mistral Small 4 Output / 1M $0.3000 $0.6000 +100.0%
Grok 4.3 Model -- added --
Command A Model -- added --
Mistral Medium 3.5 Model -- added --
Devstral Medium 2 Model -- added --
Devstral Small 2 Model -- added --
Ministral 14B Model -- added --
Mistral NeMo Model -- added --
Mixtral 8x7B Model -- added --
Mistral 7B Model -- added --
GPT-OSS 120B (Together) Model -- added --
GPT-OSS 20B (Together) Model -- added --
DeepSeek V4 Pro (Together) Model -- added --
MiniMax M2.7 (Together) Model -- added --
Kimi K2.6 (Together) Model -- added --
GLM-5.1 (Together) Model -- added --
Qwen3.6-Plus (Together) Model -- added --

(2 changes)

Model Field Old New Change
DeepSeek V4 Flash Cached Input / 1M $0.0280 $0.0028 -90.0%
DeepSeek V4 Pro Cached Input / 1M $0.1450 $0.0145 -90.0%

(1 change)

Model Field Old New Change
DeepSeek V3.1 (Together) Model -- added --

(2 changes)

Model Field Old New Change
GPT-5.5 Model -- added --
GPT-5.5 Pro Model -- added --

(4 changes)

Model Field Old New Change
DeepSeek V4 Flash Model -- added --
DeepSeek V4 Pro Model -- added --
DeepSeek V3.2 (Chat) Model removed -- --
DeepSeek V3.2 (Reasoner) Model removed -- --

(14 changes)

Model Field Old New Change
Gemini 2.5 Pro (>200k tokens) Model -- added --
Magistral Medium Model -- added --
Magistral Small Model -- added --
Mistral Large 3 Input / 1M $2.0000 $0.5000 -75.0%
Mistral Large 3 Output / 1M $6.0000 $1.5000 -75.0%
Mistral Medium 3 Model -- added --
Pixtral 12B Model -- added --
GPT-5.4 Pro Model -- added --
Sonar Model -- added --
Sonar Deep Research Model -- added --
Sonar Reasoning Pro Model -- added --
Sonar Small Online Model removed -- --
Sonar Large Online Model removed -- --
Sonar Huge Online Model removed -- --

(6 changes)

Model Field Old New Change
Grok 4 1 Fast Non Reasoning Model -- added --
Grok 4 1 Fast Reasoning Model -- added --
Llama 3.3 70b Versatile Model -- added --
Llama 3.1 8b Instant Model -- added --
Openai/gpt Oss 120b Model -- added --
Openai/gpt Oss 20b Model -- added --

(2 changes)

Model Field Old New Change
Gemini 3.1 Flash-Lite Cached Input / 1M $0.0300 $0.0250 -16.7%
Gemini 2.5 Pro Cached Input / 1M $0.1300 $0.1250 -3.8%

(22 changes)

Model Field Old New Change
Ministral 8B Model -- added --
Ministral 3B Model -- added --
Mixtral 8x22B Model -- added --
Pixtral Large Model -- added --
Command R+ 08-2024 Model -- added --
Command R 08-2024 Model -- added --
Command R7B Model -- added --
Embed v3 English Model -- added --
Embed v3 Multilingual Model -- added --
Rerank v3 Model -- added --
DeepSeek R1 (Together) Model -- added --
DeepSeek V3 (Together) Model -- added --
Llama 3.3 70B (Together) Model -- added --
Llama 3.1 405B (Together) Model -- added --
Qwen 2.5 72B (Together) Model -- added --
Mixtral 8x22B (Together) Model -- added --
Mistral Large (Together) Model -- added --
Sonar Small Online Model -- added --
Sonar Large Online Model -- added --
Sonar Huge Online Model -- added --
Sonar Pro Model -- added --
Claude Opus 4.7 Model -- added --

(32 changes)

Model Field Old New Change
Grok 3 Model removed -- --
Grok 3 Mini Model removed -- --
Grok 2 Vision Model removed -- --
Command R+ 08-2024 Model removed -- --
Command R 08-2024 Model removed -- --
Command R7B Model removed -- --
Embed v3 English Model removed -- --
Embed v3 Multilingual Model removed -- --
Rerank v3 Model removed -- --
DeepSeek R1 (Together) Model removed -- --
DeepSeek V3 (Together) Model removed -- --
Llama 3.3 70B (Together) Model removed -- --
Llama 3.1 405B (Together) Model removed -- --
Qwen 2.5 72B (Together) Model removed -- --
Mixtral 8x22B (Together) Model removed -- --
Mistral Large (Together) Model removed -- --
Sonar Small Online Model removed -- --
Sonar Large Online Model removed -- --
Sonar Huge Online Model removed -- --
Sonar Pro Model removed -- --
Ministral 8B Model removed -- --
Ministral 3B Model removed -- --
Mixtral 8x22B Model removed -- --
Pixtral Large Model removed -- --
Llama 3.3 70B Versatile Model removed -- --
Llama 3.1 8B Instant Model removed -- --
Llama 3.2 90B Vision Model removed -- --
Llama 3.2 11B Vision Model removed -- --
Mixtral 8x7B Model removed -- --
Gemma2 9B Model removed -- --
DeepSeek R1 Distill Llama 70B Model removed -- --
Qwen 2.5 32B Model removed -- --

Methodology

Pricing snapshots are taken daily at 06:00 UTC from each provider's canonical pricing page. Changes are computed by diffing consecutive daily snapshots. All prices in USD per 1 million tokens. Historical data is available programmatically via /api/price-history.json. API docs.