Model Pricing List

Detailed per-model pricing on GreenNode MaaS — Input, Output, and Cache unit prices per 1,000,000 tokens (or per image for image generation models).

The model catalog and pricing below apply under the transition starting August 3, 2026. See the timeline details in Release Notes.

1. Token-based Models

Unit: VND / 1,000,000 tokens. See how the pricing formula is applied in Pricing.

Model	Type	Old Price (VND)	New Price (VND)
Kimi K2.7 Code	Input	–	26,937
Kimi K2.7 Code	Output	–	113,420
Kimi K2.7 Code	Cache	–	5,387
Kimi 2.6	Input	36,482	26,937
Kimi 2.6	Output	153,608	113,420
Kimi 2.6	Cache	6,144	4,537
GLM-5.2	Input	–	39,697
GLM-5.2	Output	–	124,762
GLM-5.2	Cache	–	5,387
Opus 4.8	Input	–	141,775
Opus 4.8	Output	–	708,875
Opus 4.8	Cache	–	14,178
Sonnet 4.6	Input	115,206	85,065
Sonnet 4.6	Output	576,030	425,325
Sonnet 4.6	Cache	11,521	8,506
Haiku 4.5	Input	–	28,355
Haiku 4.5	Output	–	141,775
Haiku 4.5	Cache	–	2,836
Gemma 4 31B-IT	Input	7,680	5,671
Gemma 4 31B-IT	Output	19,201	14,178
Gemma 4 31B-IT	Cache	3,456	2,552
Qwen 3.7 Plus	Input	15,361	11,342
Qwen 3.7 Plus	Output	61,443	45,368
Qwen 3.7 Plus	Cache	3,072	2,268
Qwen 3.6 Plus	Input	–	14,178
Qwen 3.6 Plus	Output	–	85,065
Qwen 3.6 Plus	Cache	–	17,722
Qwen 3.6 Flash	Input	–	7,089
Qwen 3.6 Flash	Output	–	42,533
Qwen 3.6 Flash	Cache	–	532
MiniMax M3	Input	–	8,506
MiniMax M3	Output	–	34,026
MiniMax M3	Cache	–	1,701
MiniMax M2.5	Input	11,521	8,506
MiniMax M2.5	Output	46,082	34,026
MiniMax M2.5	Cache	1,152	851
GPT-5.4	Input	–	70,888
GPT-5.4	Output	–	425,325
GPT-5.4	Cache	–	7,089
GPT-5.4 Mini	Input	–	21,266
GPT-5.4 Mini	Output	–	127,598
GPT-5.4 Mini	Cache	–	2,127
GPT-5.4 Nano	Input	–	5,671
GPT-5.4 Nano	Output	–	35,444
GPT-5.4 Nano	Cache	–	567
GPT-5	Input	48,003	35,444
GPT-5	Output	384,020	283,550
GPT-5	Cache	4,800	3,544
GPT-5 Mini	Input	9,601	7,089
GPT-5 Mini	Output	76,804	56,710
GPT-5 Mini	Cache	1,152	709
GPT-5 Nano	Input	1,920	1,418
GPT-5 Nano	Output	15,361	11,342
GPT-5 Nano	Cache	384	142
GPT-4o	Input	96,005	70,888
GPT-4o	Output	384,020	283,550
GPT-4o	Cache	48,003	35,444
GPT-4o Mini	Input	5,760	4,253
GPT-4o Mini	Output	23,041	17,013
GPT-4o Mini	Cache	2,880	2,127
GPT-OSS 20B	Input	1,920	1,985
GPT-OSS 20B	Output	7,680	4,253
GPT-OSS 120B	Input	13,441	4,253
GPT-OSS 120B	Output	28,802	17,013

2. Embedding & Rerank Models

Model	Type	Old Price (VND)	New Price (VND)
Cohere Embed v4	Input	–	7,089
Cohere Rerank v4.0 pro	Input	–	71
Qwen3 Embedding 8B	Input	1,920	1,134

3. Image Generation Models (gpt-image-2)

Unit: VND / image, by resolution and quality configuration. See how image pricing is applied in Pricing.

Model — Configuration	Old Price (VND)	New Price (VND)
gpt-image-2, 1024x1024, High	8,103	5,983
gpt-image-2, 1024x1024, Medium	2,035	1,503
gpt-image-2, 1024x1024, Low	230	170
gpt-image-2, 1024x1536, High	6,336	4,679
gpt-image-2, 1024x1536, Medium	1,574	1,163
gpt-image-2, 1024x1536, Low	192	142
gpt-image-2, 1536x1024, High	6,336	4,679
gpt-image-2, 1536x1024, Medium	1,574	1,163
gpt-image-2, 1536x1024, Low	192	142

A model with no Old Price (shown as "–") is a new addition to the catalog with no prior pricing.