Gemini 4 Neon vs Mistral Large 4
Comparer des modèles
Gemini 4 Neon is more worth it — 65% vs 50%
How we score worth-it? →Google's fast Gemini 4 reasoning model for software engineering, knowledge work and science — multimodal, 1M context.
- No one could fail to love Neon. As the model that currently balances speed and quality, understands plain human speech best, speaks it best, is the most creative, the most up to date, and has the lowest moral floor, it even gives you a psychological massage while you get work done - an excellent experience.— v2ex ↗
- It's good to use. With extra turned on, this small model runs about the same as Carbon, just several times faster. It really is a finely calibrated slot - the sweet spot for daily use.— v2ex ↗
- The goat model for writing in the humanities and social sciences.— bilibili ↗
- Insane - this new model is about 1.5x slower than Opus 5.5, but the results are already very close.— bilibili ↗
- Also i am happy that the models are really good and i have no issues in their performance, very capable models keep up the good work— google-forum ↗
- Today my experience with Neon medium was really bad. It's fast, but the delivery quality is pretty poor - I had to step in and correct it several times. For now I'll obediently go back to Carbon.— v2ex ↗
- Even Gemini's Neon and Carbon have serious hallucination problems, and their coding is all surface work - it looks like the goal was achieved, but look closely at the implementation details and there are problems everywhere.— v2ex ↗
- In a single 50-minute single-threaded session using Gemini 4 Neon (Extra), Antigravity consumed ~15% of my weekly Ultra 5x quota.— google-forum ↗
- I was thinking of giving Gemini 4 Neon a try as a code reviewer instead of GPT-6.2 Sol, and found this in their FAQ: "Using third-party software, tools, or services to access Antigravity is a violation of our Terms of Service". I'd like to keep my Gmail account, so Google can keep their model and will not get my business.— hn ↗
- Well, I use my product as a benchmark for this new models and I have to say that Argon is still the most efficient model by far for my use case. Carbon, Hydrogen, and now Neon all use 3 to 4 times higher tokens compared to Argon for creating demos. Neon was the first Google model that passed 8192 MAX OUTPUT TOKEN limit, which is disappointing a little bit.— hn ↗
“127 tokens/s — roughly 3x faster than the slower members of the Gemini 4 family.”
“Reviewers call it a coding machine: strong on complex coding projects, on par with top-tier rivals.”
“Quality sits below Carbon on first hands-on runs, but it is ~4x faster and about 2x cheaper per task.”
European open-weights flagship with aggressive preview pricing.
- High one is actually much better — the feet connect to the pedals, no hub cap, relatively proper position.— hn
- Moving a lot of our stuff over to Mistral. Very much happier keeping data in Europe rather than leeching it all over the US.— hn ↗
- Mistral has been great for us as a provider for our EU-based company... since they added more features and improved their own model, we are quite happy.— hn ↗
- The dense graph pins at 16.1 GiB, and the model runs on a single 32 GB consumer card. Thank you for shipping the reference modeling code — five details are easy to get wrong and all of them change the answers.— huggingface ↗
- Mistral is good. "Only" 250 Euro, but GLM is cheap af, and mistral has absolutely zero of these weird-ass "5h limits" that became the defacto standard for some reason.— hn ↗
- Still too expensive and very slow when thinking. We need one more generation from Mistral and 50% of this already 'slashed' price to be competitive.— hn ↗
- I tried mistral 4 when it came out and it honestly felt like going back in time 6 months — my agents were making all kinds of mistakes I'd forgotten agents used to make.— hn ↗
- The benchmark numbers are dreadful and the cost is higher than Sol. There are a ton of models that are a lot smarter for a lot less money.— hn ↗
- Mistral 4 IMHO is not that good. I ran the full NVFP4 weight and couldn't stomach it after 3 days.— hn ↗
- When I last tried Mistral Large 4 for a code documentation task, it ignored my SKILL.md files and it had a few hallucinations. I wondered if I had failed to prompt it properly or if there was some problem with my harness.— hn ↗
“High-effort mode renders the pelican correctly — feet connect to pedals”
“Far from AGI — it just regenerated a well-known benchmark image”
“Aggressive preview pricing at $0.68/$2.09”
Which should you pick?
The gap is clear: Gemini 4 Neon sits at 65% while Mistral Large 4 is at 50%, a 15-point difference. If you just want the safer pick, start with Gemini 4 Neon.
On entry price, Mistral Large 4 is cheaper ($0.7 vs $3, Gemini 4 Neon costs about 329% more).
Both are in the Chat category with similar coverage; the real difference is stability and cost — run one of your own tasks through each before deciding.
What users actually said about Gemini 4 Neon: "No one could fail to love Neon. As the model that currently balances speed and quality, un…"
FAQ
Is Gemini 4 Neon or Mistral Large 4 more worth it?
Gemini 4 Neon is more worth it, with a 65% worth-it score vs 50% for Mistral Large 4.
How much does Gemini 4 Neon cost vs Mistral Large 4?
Gemini 4 Neon: $3/$15 per 1M (cached $0.15) · Mistral Large 4: $0.70/$2.10 per 1M.
What do users say about Gemini 4 Neon vs Mistral Large 4?
Gemini 4 Neon: 127 tokens/s — roughly 3x faster than the slower members of the Gemini 4 family.. Mistral Large 4: High-effort mode renders the pelican correctly — feet connect to pedals.