meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 vs Qwen: Qwen3.6 Flash
They cost about the same for input tokens: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 $0.20 per M, Qwen: Qwen3.6 Flash $0.1875 per M. meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 is 1.4× cheaper for output tokens: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 costs $0.80 per M and Qwen: Qwen3.6 Flash costs $1.125 per M. About the same: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 takes up to 1,048,576 tokens of context and Qwen: Qwen3.6 Flash up to 1,000,000. meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 accepts image, text; Qwen: Qwen3.6 Flash accepts image, text, video.
| meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 | Qwen: Qwen3.6 Flash | |
|---|---|---|
| Model id | meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 | qwen/qwen3.6-flash |
| Input | $0.20 per M | $0.1875 per M |
| Output | $0.80 per M | $1.125 per M |
| Context length | 1,048,576 tokens | 1,000,000 tokens |
| Accepts | image, text | image, text, video |
| Produces | text | text |
| Served by | deepinfra | alibaba |
Prices are the providers' list prices per million tokens, as of . Both run on your own provider key through one OpenAI-compatible endpoint, with no fee on the tokens.