French AI company Mistral AI has released the open-source model Mistral Medium3.5, but the market response has been lukewarm

BlockBeats News, May 1st: French AI startup Mistral AI released the next-generation open-source model Mistral Medium 3.5 on April 29th. The model features a dense architecture with 1.28 trillion parameters and simultaneously launched the remote coding agent tool Mistral Vibe CLI and the multi-step AI Agent feature Le Chat.

Official data shows that Mistral Medium 3.5 scored 77.6% on the SWE-Bench Verified programming benchmark test and supports unified inference and coding capabilities. However, with an API pricing of up to $1.5 per million input tokens and $7.5 per output token, significantly higher than Chinese open-source models, it has sparked widespread community scrutiny.

The current top open-source model rankings are mainly dominated by Chinese models such as Alibaba's Qwen, Zhifup GLM, and Xiaomi's MiMo-V2. Among them, Qwen 3.6 has only 270 billion parameters but has already achieved 72.4% on SWE-Bench and is completely open source and free.

However, some developers believe that Mistral's current greatest value lies in its "non-U.S., non-China" European identity. Due to compliance with GDPR and European on-premise deployment requirements, its model has been used for private deployment by large European institutions including HSBC. Market sentiment suggests that Mistral's competitiveness stems more from geopolitics and enterprise compliance rather than pure model performance.