Cheapest LLM APIs

277 models with published pricing, ranked by blended $/1M tokens. 564 models without published pricing are excluded — never shown as free.

Cheapest overall

ModelDeveloperIn $/1MOut $/1MBlendedECI
Qwen3.7 Flash Alibaba 0.03 0.13 0.055 —
Nova Micro Amazon 0.035 0.14 0.0612 —
Command R7B Cohere 0.0375 0.15 0.0656 —
Command R7B Arabic Cohere 0.0375 0.15 0.0656 —
Qwen Turbo Alibaba 0.05 0.20 0.0875 —
Ministral 3 3B Mistral 0.10 0.10 0.10 —
Nova Lite Amazon 0.06 0.24 0.105 —
Qwen-Omni Turbo Alibaba 0.07 0.27 0.12 —
GPT-5 Nano OpenAI 0.05 0.40 0.1375 139.2
Qwen Flash Alibaba 0.05 0.40 0.1375 —
Ministral 3 8B Mistral 0.15 0.15 0.15 —
Mistral Small 3.2 Mistral 0.10 0.30 0.15 —
Pixtral 12B Mistral 0.15 0.15 0.15 —
Step 3.5 Flash StepFun 0.1046 0.3138 0.1569 —
Step 3.5 Flash 2603 StepFun 0.1046 0.3138 0.1569 —
Mercury 3 Inception 0.10 0.35 0.1625 —
GLM-4.7-FlashX Zhipu AI 0.0747 0.4482 0.1681 —
GPT-4.1 nano OpenAI 0.10 0.40 0.175 127.0
Gemini 2.5 Flash-Lite Google 0.10 0.40 0.175 —
MiMo-V2.5 Xiaomi 0.14 0.28 0.175 —
MiMo-V2.6-Flash Xiaomi 0.14 0.28 0.175 —
Qwen3.5 Flash Alibaba 0.10 0.40 0.175 —
Solar Mini 4 Upstage 0.10 0.40 0.175 —
Voxtral Small (latest) Mistral 0.10 0.40 0.175 —
GLM-5.3-Flash Zhipu AI 0.1195 0.4183 0.1942 153.6
Viv Fast Vivgrid 0.13 0.40 0.1975 —
Claude Haiku 5.5 Anthropic 0.10 0.50 0.20 160.2
GPT-6 Luna OpenAI 0.10 0.50 0.20 156.7
Ministral 3 14B Mistral 0.20 0.20 0.20 —
GLM-4.6V Zhipu AI 0.1494 0.4482 0.2241 —
Qwen3.8 Flash Alibaba 0.15 0.47 0.23 —
Qwen3.8 Omni Flash Alibaba 0.15 0.47 0.23 —
Jamba Mini AI21 Labs 0.20 0.40 0.25 —
Solar Pro 2 Upstage 0.25 0.25 0.25 —
Solar Pro 3 Upstage 0.25 0.25 0.25 —
Command R Cohere 0.15 0.60 0.2625 —
DeepSeek V4.1 Flash 1017 DeepSeek 0.15 0.60 0.2625 159.3
GPT-4o mini OpenAI 0.15 0.60 0.2625 124.0
MiniMax-M3.1-Flash MiniMax 0.15 0.60 0.2625 —
Mistral Small (latest) Mistral 0.15 0.60 0.2625 —
Mistral Small 4 Mistral 0.15 0.60 0.2625 —
Sarvam 105B Sarvam AI 0.1579 0.6316 0.2763 —
Qwen3 8B Alibaba 0.18 0.70 0.31 132.2
GLM-4.5-Air Zhipu AI 0.1195 0.8964 0.3137 —
DeepSeek Chat DeepSeek 0.28 0.42 0.315 —
DeepSeek Reasoner DeepSeek 0.28 0.42 0.315 —
Qwen-VL Plus Alibaba 0.21 0.63 0.315 —
Mercury 2.5 Inception 0.20 0.75 0.3375 —
Mercury 2 Inception 0.25 0.75 0.375 —
Mercury Edit 2 Inception 0.25 0.75 0.375 —

Cheapest at each capability level

Bands are Epoch Capabilities Index (ECI) quartiles computed from the priced, ECI-scored population itself, not fixed numbers: Top quartile (ECI ≥ 158.3), Upper-middle quartile (ECI 154.0–158.3), Lower-middle quartile (ECI 146.5–154.0), Bottom quartile (ECI < 146.5). Top 5 cheapest per band, by blended price.

Top quartile (ECI ≥ 158.3)

ModelDeveloperECIBlended $/1M
Claude Haiku 5.5 Anthropic 160.2 0.20
DeepSeek V4.1 Flash 1017 DeepSeek 159.3 0.2625
Gemini 3.7 Flash Google 158.4 1.50
Gemini 3.8 Flash Google 158.6 1.50
GPT-6.1 Terra OpenAI 167.8 2.00

Upper-middle quartile (ECI 154.0–158.3)

ModelDeveloperECIBlended $/1M
GPT-6 Luna OpenAI 156.7 0.20
GPT-5.6 Luna OpenAI 156.8 0.45
DeepSeek V4 Pro 0813 DeepSeek 155.9 0.99
Mistral Large 4 Mistral 154.7 1.05
Gemini 3.6 Flash Google 155.2 1.50

Lower-middle quartile (ECI 146.5–154.0)

ModelDeveloperECIBlended $/1M
GLM-5.3-Flash Zhipu AI 153.6 0.1942
DeepSeek V4 Flash DeepSeek 146.9 0.525
MiniMax-M3 MiniMax 147.6 0.525
Qwen3.7 Plus Alibaba 147.8 0.70
Kimi K2.5 Moonshot AI 148.7 1.20

Bottom quartile (ECI < 146.5)

ModelDeveloperECIBlended $/1M
GPT-5 Nano OpenAI 139.2 0.1375
GPT-4.1 nano OpenAI 127.0 0.175
GPT-4o mini OpenAI 124.0 0.2625
Qwen3 8B Alibaba 132.2 0.31
GPT-5.4 nano OpenAI 146.4 0.4625

Cheapest input price

ModelDeveloperIn $/1M
Qwen3.7 FlashAlibaba0.03
Nova MicroAmazon0.035
Command R7BCohere0.0375
Command R7B ArabicCohere0.0375
GPT-5 NanoOpenAI0.05
Qwen FlashAlibaba0.05
Qwen TurboAlibaba0.05
Nova LiteAmazon0.06
Qwen-Omni TurboAlibaba0.07
GLM-4.7-FlashXZhipu AI0.0747

Cheapest output price

ModelDeveloperOut $/1M
Ministral 3 3BMistral0.10
Qwen3.7 FlashAlibaba0.13
Nova MicroAmazon0.14
Command R7BCohere0.15
Command R7B ArabicCohere0.15
Ministral 3 8BMistral0.15
Pixtral 12BMistral0.15
Ministral 3 14BMistral0.20
Qwen TurboAlibaba0.20
Nova LiteAmazon0.24

FAQ

How is 'cheapest' determined?

Models are ranked by blended $/1M tokens (official pricing from models.dev/LiteLLM). Only models with published pricing are ranked — 564 models without published pricing are excluded, never assumed free.

What is the cheapest LLM API right now?

Qwen3.7 Flash by Alibaba, at $0.055/1M blended tokens, per the latest sync.

Why does 'cheapest at each capability level' matter?

The single cheapest model overall may be too weak for your task. Grouping by Epoch Capabilities Index (ECI) quartile surfaces the cheapest option that still clears a given capability bar, instead of overpaying for capability you don't need.

Are these prices guaranteed accurate?

Prices are official list prices as published by models.dev/LiteLLM, refreshed daily. Always confirm current pricing with the provider before committing production spend — and check the conflict/off-peak notes where flagged.

Want capability instead of raw price? See the best-value leaderboard (ECI per dollar), or the full ECI leaderboard.