リリース2026/09
生成速度0 トークン/秒
初動レイテンシ (TTFT)0.00s
入力料金$0.05/1M
出力料金$0.15/1M
ランキング
レビュー
Agnes 3.0 Flash excels at reasoning and solving complex problems, making it a strong fit for users who need high accuracy in research or programming. However, if the primary goal is casual conversation or deploying autonomous AI agents, other models may be better suited. Ideal for specialists and developers looking to leverage deep reasoning.
長所
- Outstanding reasoning performance, ranking #10 on the AIM leaderboard
- Strong capability in solving complex logic and scientific problems
- High benchmark scores on GPQA Diamond, delivering reliable output in deep reasoning tasks
短所
- Average general chat performance, ranking #28 on AIM
- Weak agentic capabilities (#32), limiting its effectiveness in multi-step autonomous workflows
- Higher cost and latency compared to models optimized for fast conversational chat
ユースケース
Data analysis and scientific inference for researchSolving advanced mathematical, logical, or specialized technical problemsAssisting with code generation and review, especially in projects requiring rigorous reasoning