Budget Models

Cheapest LLM APIs in 2026: Under $1 per Million Tokens

Looking for a cheap LLM API? These models cost less than $1 per million input tokens while still delivering solid performance. Perfect for high-volume applications, rapid prototyping, batch processing, and anyone watching their API bill. Sorted by price so you can find the cheapest option that meets your quality needs.

222 models·47 providers·$0.0100 to $0.9000/1M input

Top Picks

Compare Models

vs
Compare

Popular comparisons

Related Categories

Cost per 1M Tokens / Flagship Models

Input
Output

All Budget Models

222 models
ProviderModelInput $/1MOutput $/1MContextQuality
LiquidLfm2 8b A1b$0.0100$0.020032.8KN/A
LiquidLfm 2.2 6b$0.0100$0.020032.8KN/A
Ibm-graniteGranite 4.0 H Micro$0.0170$0.1100131KN/A
GoogleGemma 3n E4b It$0.0200$0.040032.8K
65.8Frontier
MetaLlama Guard 3 8b$0.0200$0.0600131.1KN/A
MetaLlama 3.1 8b Instruct$0.0200$0.050016.4K
46.6Good
MistralMistral Nemo$0.0200$0.0400131.1KN/A
MetaLlama 3.2 1b Instruct$0.0270$0.200060K
28.7Economy
LiquidLfm 2 24b A2b$0.0300$0.120032.8KN/A
OpenAIGPT Oss 20b$0.0300$0.1100131.1K
65.7Frontier
QwenQwen2.5 Coder 7b Instruct$0.0300$0.090032.8KN/A
MistralMistral Small 3.1 24b Instruct$0.0300$0.1100131.1K
63.0Strong
GoogleGemma 2 9b It$0.0300$0.09008.2K
58.8Strong
MetaLlama 3 8b Instruct$0.0300$0.04008.2K
48.7Good
QwenQwen Turbo$0.0325$0.1300131.1KN/A
AmazonNova Micro V1$0.0350$0.1400128KN/A
CohereCommand R7b 12 2024$0.0375$0.1500128KN/A
OpenAIGPT Oss 120b$0.0390$0.1900131.1K
47.7Good
NVIDIANemotron Nano 9b V2$0.0400$0.1600131.1KN/A
GoogleGemma 3 4b It$0.0400$0.0800131.1K
63.0Strong
GoogleGemma 3 12b It$0.0400$0.1300131.1K
69.9Frontier
QwenQwen 2.5 7b Instruct$0.0400$0.100032.8KN/A
Sao10kL3 Lunaris 8b$0.0400$0.05008.2KN/A
Arcee-aiTrinity Mini$0.0450$0.1500131.1KN/A
MetaLlama 3.2 11b Vision Instruct$0.0490$0.0490131.1KN/A
QwenQwen3.5 9b$0.0500$0.1500256KN/A
NVIDIANemotron 3 Nano 30b A3b$0.0500$0.2000262.1K
65.7Frontier
OpenAIGPT 5 Nano$0.0500$0.4000400K
49.9Good
QwenQwen3 8b$0.0500$0.400041.0KN/A
AllenaiOlmo 2 0325 32b Instruct$0.0500$0.2000128K
53.8Strong
MistralMistral Small 24b Instruct 2501$0.0500$0.080032.8K
57.8Strong
MetaLlama 3.2 3b Instruct$0.0510$0.340080K
38.6Good
xAIGlm 4.7 Flash$0.0600$0.4000202.8K
74.7Frontier
QwenQwen3 14b$0.0600$0.240041.0KN/A
AmazonNova Lite V1$0.0600$0.2400300KN/A
GrypheMythomax L2 13b$0.0600$0.06004.1KN/A
QwenQwen3.5 Flash 02 23$0.0650$0.26001M
80.6Frontier
MicrosoftPhi 4$0.0650$0.140016.4K
54.6Strong
BaiduErnie 4.5 21b A3b Thinking$0.0700$0.2800131.1KN/A
BaiduErnie 4.5 21b A3b$0.0700$0.2800120KN/A
QwenQwen3 Coder 30b A3b Instruct$0.0700$0.2700160KN/A
QwenQwen3 235b A22b 2507$0.0710$0.1000262.1K
52.2Strong
Bytedance-seedSeed 1.6 Flash$0.0750$0.3000262.1KN/A
OpenAIGPT Oss Safeguard 20b$0.0750$0.3000131.1KN/A
MistralMistral Small 3.2 24b Instruct$0.0750$0.2000128KN/A
GoogleGemini 2.0 Flash Lite 001$0.0750$0.30001.0M
72.0Frontier
QwenQwen3 Vl 8b Instruct$0.0800$0.5000131.1KN/A
QwenQwen3 30b A3b Thinking 2507$0.0800$0.4000131.1K
38.2Good
QwenQwen3 30b A3b$0.0800$0.280041.0K
38.2Good
QwenQwen3 32b$0.0800$0.240041.0K
42.2Good

Page 1 of 5

Deploy your AI agent for $33/mo flat.

Managed Telegram bot hosting. We handle the infrastructure.

Any model. Any channel. Zero infrastructure.