Day 7 · 2026-10-31
18 launches checked71 claims quoted word for word121 user reports linked10 verified fixes linked26 corrections publishedLive 8 days5 of 5 founding sponsor spots left
Day 7Launch day vs. day 7
AI launches, checked against real usersLast re-stamped 31 Oct 2026

AI models / Compare

EmbeddingGemma 2 vs Mistral Large 4: which delivered after 7 days?

EmbeddingGemma 2 is stamped Delivered; Mistral Large 4 is still Too early to call, so this comparison isn't settled yet.

EmbeddingGemma 2Mistral Large 4
VerdictDeliveredToo early
Claims held31
Claims mixed20
Claims failed00
Claims unclear02
User reports163
Launched6 Oct 202629 Oct 2026
Last stamped30 Oct 202631 Oct 2026

No. 009 · Google

EmbeddingGemma 2

DeliveredDay 7 verdict

The open weights, shared multimodal space and smaller sizes check out in independent tests; multilingual parity is shaky outside English, and runtime support arrived late.

Launched
6 Oct 2026
Stamped
30 Oct 2026
Claims
✓3 held · ~2 mixed · ✗0 failed · ?0 unclear
Reports
16 (7 praise · 5 problem · 4 mixed)

What they claimed

“Today, we’re launching EmbeddingGemma 2, expanding beyond text to unify code, images, video, and audio in a shared embedding space.”
✓ Holds
“Built on the Gemma 4 architecture and released under a commercially permissive Apache 2.0 license, EmbeddingGemma 2 has 740 million parameters, making it optimal for on-device inference.”
✓ Holds
“Using Matryoshka Representation Learning (MRL), developers can dynamically truncate output vectors from 768 dimensions down to 512, 256, or 128 dimensions.”
✓ Holds
“EmbeddingGemma 2 matches the strong multilingual text performance of EmbeddingGemma while delivering a significant 9.92-point improvement on code performance (in MTEB Code, from 68.76 to 78.68), making it well-suited for local codebase indexing, semantic code search, and coding agent retrieval.”
~ Mixed
“We worked closely with the following partners to ensure EmbeddingGemma 2 works immediately where you build:”
~ Mixed

Full receipt

No. 017 · Mistral AI

Mistral Large 4

Day 2 of 7
Too earlyDay 7 verdict

Weights are out, but few hands-on reports yet; the cybersecurity claim leans on one index and is untested by users.

Launched
29 Oct 2026
Stamped
31 Oct 2026
Claims
✓1 held · ~0 mixed · ✗0 failed · ?2 unclear
Reports
3 (0 praise · 3 problem · 0 mixed)

What they claimed

“ML4 is one of the world's strongest AI models for cybersecurity.”
? Unclear
“We will release the weights by the end of the month.”
✓ Holds
“significantly outperforming any open-weight model developed in the US or Europe”
? Unclear

Full receipt

Both stamps come from the public user reports linked on each receipt, under the same method. Verdicts are opinion and are never paid for.

Keep comparingall AI models