Jake (@JakeToepel) header
Jake (@JakeToepel) profile picture

Partner @ LeanScale | @ChelseaFC | Go-To-Market Guru | @TheSunDevils | @Patriots | @AZCardinals | @PHXRisingFC | @Suns | 1/10th of L4J | Bad at Golf

Phoenix, AZ
Joined August 2015
Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)Photo by Jake (@JakeToepel)
Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding. • Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure. • Available to all via API today. Working with cybersecurity partners privately. Open weights release end of October.
🤖 Made with AI
2,029
4,551
2,814
40,572
5,777,386
Chelsea beating a cheating Man City in a UCL final doesn’t get spoken about enough
21
946
34
10,427
122,677
NETO! 🔥 🔵 3-2 🟡 [60] #CFC | #CarabaoCup
210
518
83
4,472
250,178
Kit’s employee benefits are S-tier
84
34
30
2,811
708,903
Jake retweeted
God was the first vibe coder. “Let there be light.” No design doc. No tests. No staging environment. Just shipped the universe directly to production.
434
2,659
306
25,481
827,298
“we’ve got over 50 beers on tap” wow that’s great miller lite please
249
2,986
228
66,560
3,081,699
Jake retweeted
Odyssey plot hole: Why was the Cyclops making goats cheese with no profit incentive or shareholders to create value for?
277
8,370
274
126,317
2,501,343
JUST IN: Gen Z electricians working on AI data centers are reportedly earning up to $280,000 a year, as demand surges.
411
1,107
221
19,085
1,708,579
Replying to @Breaking911
TERRIBLE. Permanent DST means 8:45am sunrise in most states. Tell the Senate to amend to permanent Standard Time!
369
491
84
3,887
559,274
I've seen projects that spend 3-5x more energy in alignment and exec reviews than engineering. With coding agents, we exponentially reduce the time spent in engineering and increase the effort required to align vision, strategy, timelines, etc.
5
6
1
65
6,540
This is mad btw 😂
June 29, 2000: Patrick Kluivert misses a PK against Italy, hitting the left upright. June 29, 2026: Justin Kluivert misses a PK against Morocco, hitting the left upright.
28
1,678
5
26,508
1,928,686
225
6,337
170
125,720
4,259,715
BREAKING: Landon Hairston, a Golden Spikes Finalist and National Player of the Year by Baseball America, just announced on MLB Network that he will return to Arizona State for his junior season. 247sports.com/college/arizon…
29
121
34
873
106,880
The CDC is seeking volunteers to help screen for Ebola in US airports. We had a good run.
The US Centers for Disease Control and Prevention is seeking volunteers from its workforce to go to domestic airports and help screen for Ebola as the deadly viral outbreak in the Democratic Republic of Congo and Uganda intensifies bloomberg.com/news/articles/…
165
1,555
51
8,169
887,339
140
3,901
229
51,453
1,285,262
Jake retweeted
39
7,346
91
109,106
1,010,745
Jake retweeted
Breaking: Pep Guardiola has joined Anthropic
Pep Guardiola to step down after incredible decade as City Manager 🩵 🔗 mancity.co/Pep-Guardiola
67
486
43
9,043
891,627
Jake retweeted
Introducing Hanta Virus
Branding Hanta virus
451
3,974
430
43,537
3,423,792
Mistral Medium 3.5 is interesting less for the benchmarks and more for the positioning. Look at who they're comparing against: Kimi, Qwen, GLM, Claude (Sonnet). Not GPT, not Gemini. And i dont mean that in a negative way! With Aleph Alpha being acquired by Cohere last week, Mistral is now the only non-US, non-Chinese lab still in the frontier conversation. At 128B dense with open weights, they're making a different bet than the Chinese MoE models in that chart (which activate only 17-40B params despite being 400B-1T total). Mistral is trading inference efficiency for consistency. The Collie score (95.8, best in class by a wide margin) tells you where they're aiming: not raw reasoning, but the most reliable model to actually follow instructions in production. That's a European enterprise pitch, not a benchmark race. Very solid release from Mistral!
Replying to @mistralvibe
Mistral Medium 3.5, a new flagship model in public preview by @MistralAI that merges instruction-following, reasoning, and coding into a single 128B dense model with a 256k context window and configurable reasoning effort. It's a new default model for Mistral Vibe and Le Chat. Released as open weights, under a modified MIT license.
44
59
8
608
64,469
Jake retweeted
me and Claude after we took down prod together
me and him after the dumbest fight but we can’t stay mad at each other
44
1,160
15
19,402
835,907