Mistral Medium 3.5

“Mistral large 3/medium 3.5/small 4 have all been pretty dated for a while”

“Mistral's medium is absolutely good enough to help with a ton of coding tasks.”

“their Mistral Medium 3.5 with reasoning is worse than Gemma 4 31B and Glimmer 30B”

“asking Mistral medium to rewrite all of”

“The only solution I’ve found that works is asking Mistral medium to rewrite all of Claude’s documentation”

“twice the European counterpart (albeit dated for today's standards?)”

“The new Mistral Medium 3.5 is also a big improvement over devstral-2”

“considerably better at writing than Opus”

“and Mistral Medium down to some Qwen 3.6 35B for effectively peanuts compared to Anthropic.”

“Mistral Medium 3.5 being significantly better at real-time decision making than they are at coding”

“perfectly fine if you're willing to talk to it in a more fine-grained way about actual code”

“new Medium 3.5 is far more expensive than previous Mistral models”

“It will probably be enough for plenty of tasks, though there's also a noticeable difference between that and up to 1M.”

“lower sized models, lower costs, overall good enough performance”

“Being near the Pareto frontier of inference cost vs. output quality.”

“But what is the rationale for running a dumb model? Because it can ocasionally produce something passable?”

“This is going to be much slower.”

“considering the model is getting a bit old, I wouldn't expect them to keep offering it forever”

“releasing a model today that's only on par with open-source dense models a quarter of the size is kind of a flop”

“Not really. The benchmarks are based on F8_E4M3 and you’re not running that on any Mac.”

“it's working great for my usual test prompts”

“For it's size, that's really good!”

“Mistral continuing to ship credible models is good for the market.”

“When I try other html / js prompts it also lacks behind china models from over half a year ago.”

“there is a significant difference between Mistral and Kimi or Gemini”

“Not sure it will beat Sonet at Q4.”

“it's open weight and is tiny compared to a lot of it's competitors”

“Given what Vibe already did in the previous versions with codestral-v2, that's great news.”

“the fact the Gemma 4 26B A4B I'm running locally can blow it out of the water doesn't give me high confidence for the model”

“This model is a very good option for enterprise self host.”

“It's giving me better results than mistral-large for my use cases.”

“most aggressive open-weight pricing I've seen anywhere near frontier capability”

“This beats the latest Sonnet while running locally”

“This particular Mistral Medium looks to be uncompetitive at that pricing.”

“Unfortunately they only compare to old “all other open models”. There are probably over 10 other open models better than it by now.”

“model and national diversity is great. This looks like a solid foundation to build on”

“within a few percent of SOTA that might be good enough for a lot of people”

“It's SWE-Bench Verified. A benchmark Open-AI stopped using two months ago due to contamination.”

“I use mistral-medium-3.1 for a lot of random daily tasks”

“It's been insanely fast, cheap, reliable, and follows formatting instructions to the letter.”