Unverified50% confidenceFactExact time
The parameter counts published for Decision 3.0 include the vision encoder, with the vision encoder accounting for 0.33B in the 4B version and 0.46B in the 27B version.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe 27B version of Decision 3.0 scored 64.7 on the text leaderboard and 71.6 on the vision leaderboard, leading all participants on both leaderboards.68% similarUnverifiedMistral Large 4's vision encoder has been reduced from 2.5 billion parameters in the previous generation to 1.6 billion parameters.67% similarUnverifiedThe 9B version of Decision 3.0 (codenamed d3-flash) outperformed the previous generation's 27B model, Decision 2.0, in evaluations.66% similarUnverifiedNo Decision 3.0 model was exposed to video data during training, yet on all 19,140 questions in the Perception Test validation set, the 27B version achieved a zero-shot score of 77.4 and the 4B version scored 70.3.65% similarUnverifiedStrata automatically recommends quantization tiers based on memory: Q2 for 37.6 GB, iQ2XS (officially recommended) for close to 40 GB, higher tiers for 64 GB and above, and a Coder IQ1 version for coding scenarios.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1003459API
curl https://kongchang.com/api/v1/knowledge/claims/1003459MCP
get_claim(id=1003459)