Cheapest Language Models in 2026

Updated 7 October 2026

437 on record. Sorted by input price; every figure comes from the OpenRouter model catalogue, with the page it was read from on each entry.

437listed

#	Model	Context window	Input / 1M	Output / 1M
101	DeepSeek Pro LatestDeepSeektext	1M	$0.132	$0.396
102	Hunyuan A13B InstructTencentopen weightstext	131K	$0.141	$0.571
103	MiMo-V2.5Xiaomiopen weightstext, audio, image, video	1.1M	$0.141	$0.281
104	MiMo-V2.6-FlashXiaomiopen weightstext, image, video, audio	1.1M	$0.141	$0.281
105	Codestral 2508 (batch)Mistraltext, file	256K	$0.151	$0.451
106	Command R (08-2024)Coheretext	128K	$0.151	$0.61
107	Gemini 2.5 Flash (batch)Googlefile, image, text, audio, video	1M	$0.151	$1.251
108	Gemini 3.5 Flash Lite (batch)Googletext, image, video, file, audio	1M	$0.151	$1.251
109	GLM 5.3 FlashZ.aiopen weightstext, image, video	1M	$0.151	$0.51
110	GPT-4o-miniOpenAItext, image, file	128K	$0.151	$0.61
111	GPT-4o-mini (2024-07-18)OpenAItext, image, file	128K	$0.151	$0.61
112	Ministral 3 8B 2512Mistralopen weightstext, image	262K	$0.151	$0.151
113	Mistral Small 4Mistralopen weightstext, image	262K	$0.151	$0.61
114	Perceptron Mk1Perceptrontext, image, video	33K	$0.151	$1.51
115	Perceptron Mk1.5Perceptrontext, image, video, audio	37K	$0.151	$1.51
116	Qwen3.5-35B-A3BQwenopen weightstext, image, video	262K	$0.151	$11
117	Qwen3.6 35B A3BQwenopen weightstext, image, video	262K	$0.151	$11
118	Qwen3.8 FlashQwenopen weightstext, image, video	1M	$0.151	$0.471
119	Qwen3.8 Omni FlashQwentext, image, audio, video	1M	$0.151	$0.471
120	Qwen3 Next 80B A3B ThinkingQwenopen weightstext	262K	$0.151	$1.21
121	Qwen3 VL 30B A3B InstructQwenopen weightstext, image	262K	$0.151	$0.61
122	Solar Pro 3Upstagetext	131K	$0.151	$0.61
123	Hy3 previewTencentopen weightstext	262K	$0.181	$0.61
124	Llama Guard 4 12BMetaopen weightsimage, text	164K	$0.181	$0.181
125	Qwen3 VL 8B ThinkingQwenopen weightsimage, text	131K	$0.181	$2.11

Context bars share one log scale, 1K to 2M tokens. Prices are what each provider publishes per million tokens; 'varies' is a router that bills the model it picks.

More in Models

All models lists

Longest-context models 466 Open-weight models 178 Vision language models 295 Free models 22 Newest models 466 Reasoning models 332 Models with tool calling 395 Image generation models 9