meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 vs Qwen: Qwen3.8 Flash
Qwen: Qwen3.8 Flash is 1.3× cheaper for input tokens: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 costs $0.20 per M and Qwen: Qwen3.8 Flash costs $0.15 per M. Qwen: Qwen3.8 Flash is 1.7× cheaper for output tokens: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 costs $0.80 per M and Qwen: Qwen3.8 Flash costs $0.47 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: Qwen3.8 Flash up to 1,000,000. meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 accepts image, text; Qwen: Qwen3.8 Flash accepts image, text, video.
| meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 | Qwen: Qwen3.8 Flash | |
|---|---|---|
| Model id | meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 | qwen/qwen3.8-flash |
| Input | $0.20 per M | $0.15 per M |
| Output | $0.80 per M | $0.47 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.