Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

DeepSeek-V4-Flash Update

api-docs.deepseek.com

Post reply on HN
Print view View original

151–160 of 362 posts

Re: DeepSeek-V4-Flash Update

#151

Post by f311a »

amelius wrote: post #134

Note: if you are having success with a model, then please post what you are using it for. Writing HTML/CSS is very different from writing Rust/C++ or doing maths.

I do Python, Go and Rust. Go works the best, I would say Python is worse than Go.

Rust works perfectly fine, but when I use Rust, I usually pay closer attention to performance, so I have to guide it a bit, to improve cache locality, use simd, avoid unnecessary allocations and so on. Terra has the same issues with Rust. If you always prompt models to achieve the best performance, the code is usually no the one that I want, they optimize unnecessary/cold parts or blindly optimize stuff where compiler takes care of the optimizations already.

See my other comment for more information on how I work with it. In short, keep the changes under 1k lines, context under 120k (ask it to use subagents), drive the architecture yourself.

Re: DeepSeek-V4-Flash Update

#152

Post by ilaksh »

KaseyKim wrote: post #103
NitpickLawyer wrote: post #3

This is more exciting than k3, IMO. Dsv4 models are extremely cheap to serve. Improving their capabilities has lots of downstream effects, as it becomes "good enough" for more and more tasks. DS was serving the pro version at extremely low prices for a long time, and they've had integrations with opencode & other providers, so they likely gathered a lot of data from real developers doing real tasks (on openrouter the…

hope that deepseek become better

Have you tried the one that was just released?

Re: DeepSeek-V4-Flash Update

#153

Post by Tepix »

dnhkng wrote: post #11
NitpickLawyer wrote: post #3

This is more exciting than k3, IMO. Dsv4 models are extremely cheap to serve. Improving their capabilities has lots of downstream effects, as it becomes "good enough" for more and more tasks. DS was serving the pro version at extremely low prices for a long time, and they've had integrations with opencode & other providers, so they likely gathered a lot of data from real developers doing real tasks (on openrouter the…

Totally! This with DwarfStar delivers usable local AI (I hope!)

What are your goalposts? Depending on your requirements, there have been many moments of usable local AI. More recent ones were gpt-oss 120b and Qwen 3.6 27b.

Re: DeepSeek-V4-Flash Update

#154

Post by kzrdude »

gpugreg wrote: post #136
kzrdude wrote: post #102

Earlier quoted context omitted.

Opencode-go gives you $60 worth of DS V4 api usage for $10 per month. Right now I think it's hard to exhaust that when using flash exclusively, and plain API use might even be cheaper! Anyway, for DS usage it's a good deal.

To add to this, the $60 only applies to DeepSeek-V4-Flash and a few other models. For DeepSeek-V4-Pro, the amount is $15. https://opencode.ai/docs/go/#usage-limits Previously, OpenCode Go had higher API prices for some models, but now they lowered the API price and simultaneously reduced the allowance.

Thanks. These things change day by day I guess, AI is just moving fast (and I'm on vacation).

GPT 5.6 Luna is a new model in Go since I last checked, for example.

Re: DeepSeek-V4-Flash Update

#155

Post by benjiro29 »

PhilippGille wrote: post #122
try-working wrote: post #33

Earlier quoted context omitted.

DeepSeek themselves called it `deepseek/deepseek-v4-flash`. Pro is still like that.

Yes that's my point. The old and the new version are different in capabilities, but now when someone talks about DeepSeek V4 Flash (in benchmarks, on inference providers), you don't know which exact version it's about. Some providers like OpenRouter now call it `deepseek-v4-flash-0731`, but even in places like here on HackerNews people say things like "Sonnet is better than DeepSeek" without specifying a version or a…

I think it does not matter. Most people who use these types of models are into the IT world, and will know/be informed very fast that there is a difference. People will likely also use the 0731 behind it.

The only issue i see, is 3th party providers that have not yet updated. But that is going to be a short time periode. There is no reason to not update.

Re: DeepSeek-V4-Flash Update

#157

Post by benjiro29 »

arjie wrote: post #32
spwa4 wrote: post #4

In case people want to run it, it's DeepSeek-V4-Flash-284B-A13B. So it should just barely run on a single B300, and it's small enough that it'll barely run on an M5 Max too.

Not yet, right? That's the old DS V4 preview release. We're still waiting for the weights to come out.

Probably the same. When the same base model is trained, the weight do not tend to change a lot. GLM 5.0 > 5.1 > 5.2 are the same base model, that just kept being trained. Weights hardly change as a result. Think in the like few percentage points size difference.

Re: DeepSeek-V4-Flash Update

#158

Post by lostmsu »

ggcr wrote: post #18

Woah, a 200B model competing with GLM-5.2 and getting close to Opus 4.8. Quite impressive. If those numbers translate well to its general capabilities, with the great caching DeepSeek has, I feel like this model will get tons of usage.

Not just 200B model, it is only 160GiB.

Re: DeepSeek-V4-Flash Update

#159

Post by kzrdude »

gpugreg wrote: post #139
Lalabadie wrote: post #132

Earlier quoted context omitted.

Opencode also have a ZDR (zero data retention) deal with them – if I recall correctly, that's not something you can enable as an individual DeepSeek subscriber.

Unfortunately, all mentions of ZDR have silently been removed from the OpenCode Go page today.

Thanks, any update here is important. I use them because of good data policies..

I still find this today:

> The plan is designed primarily for international users and provides stable global access. Your data will not be used for model training.

Re: DeepSeek-V4-Flash Update

#160

Post by benjiro29 »

villish wrote: post #20
dnhkng wrote: post #2

DeepSeek V4 Flash (Preview → 2026-07-31) • Terminal Bench: 56.9 → 82.7 (+25.8) • Toolathlon: 51.8 → 70.3 (+18.5) Compared to GPT-5.6 Terra: • Terminal Bench: Flash 82.7 vs Terra 78.4 • Toolathlon: Flash 70.3 vs Terra 53.1 • DeepSWE: Flash 54.4 vs Terra 69.6 • Agents' Last Exam: Flash 25.2 vs Terra 50.4 Trading blows with Terra, which is pretty interesting. No clear winner on these benchmarks, and wildy differeing sco…

> Terminal Bench: Flash 82.7 vs Terra 78.4 Terra 87.4 https://openai.com/index/gpt-5-6/

https://www.tbench.ai/leaderboard/terminal-bench/2.1

> 78.4

The real score is always the official benchmark.

We need to see later if DS4 flash 0731 is going to maintain the score but we need to look at the official benchmarks.

Already seen a PR for DeepSWE to update the benchmark with 0731, so we can verify claimed vs official.

Post reply on HN