GitHub[208K]试用 OpenCode

CompareAI models

  1. Qwen3.8 Omni FlashAlibaba
  2. DeepSeek V4.1 FlashDeepSeek
ReasoningCodingCost efficiencyContext windowMultimodalTool use
Normalized model capability scores
ModelReasoningCodingCost efficiencyContext windowMultimodalTool use
Qwen3.8 Omni Flash100/100No data76/10081/10067/100No data
DeepSeek V4.1 Flash100/100No data68/10081/10033/100No data
Overview
Author
Context length
1M
1M
Reasoning
True
True
Input modalities
Text, Image, Audio, Video
Text, Image
Output modalities
Text
Text
Providers
Pricing
Input
$0.15/ 1M
$0.21/ 1M
Output
$0.47/ 1M
$0.84/ 1M
Cached input
$0.02/ 1M
$0.0042/ 1M
MomentumLast 2 mo
Unique users
No usage
439K
Completed sessions
No usage
7,390,118
Token share
No usage
8.3%
Tokens
No usage
55T+100%
No trend data
JUL 25SEP 18
RetentionWeek 1
Returning users
Pending
Pending

Related comparisons. Other model pairs to check.