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

DeepSeek-v4-flash-vision-exp

api-docs.deepseek.com

Post reply on HN
Print view View original

71–80 of 169 posts

Re: DeepSeek-v4-flash-vision-exp

#71

Post by mavamaarten »

LorenDB wrote: post #2

I've heard that DeepSeek v4 Flash 0731 has frequently assumed that it has vision capabilities and then resorts to inventing text-based image analysis tools when it finds that it actually can't see. In that case, this is a great upgrade for the model. Anecdotally, I had to tell 0731 to refrain from viewing screenshots since it kept breaking its sessions by trying to read images.

Yeah I've seen it a lot. It goes through the effort, unasked, of pulling screenshots off a connected device and then it's like... Oh shit yeah I can't see.

Re: DeepSeek-v4-flash-vision-exp

#72

Post by knollimar »

johndough wrote: post #17
knollimar wrote: post #5

Earlier quoted context omitted.

Oof 800 by 800 kills a lot of use cases

Might still be fine. The most recent crop of vLLMs proactively use whichever programs are available on the system (e.g. ImageMagick or PIL) to "zoom in" by cropping subimages if they can't quite make out the details.

Downsizing a higher res image to lower res means the zoom will be blurry.

Re: DeepSeek-v4-flash-vision-exp

#74

Post by leumon »

It fails the simple clock test for me which Qwen3.8 27B got (nearly) right. given an image of a clock https://files.catbox.moe/kgwa5e.png

I asked it "what time does the clock show?" (both on reasoning: high)

DS answered: The clock shows *5:10* (and 45 seconds). Here is the breakdown: * *Hour hand (red, shortest):* Pointing at the *5*. * *Minute hand (green, longest):* Pointing at the *2*, which represents 10 minutes. * *Second hand (blue, medium):* Pointing at the *9*, which represents 45 seconds.

Qwen answered: The clock shows *8:10* (with the red second hand on the 5, i.e. *8:10:25*).

- *Hour hand* (short, blue) → 8 - *Minute hand* (long, green) → 2 (10 minutes) - *Second hand* (thin, red) → 5 (25 seconds)

Correct answer is 08:09:25.

Re: DeepSeek-v4-flash-vision-exp

#75

Post by ComputerGuru »

leumon wrote: post #74

It fails the simple clock test for me which Qwen3.8 27B got (nearly) right. given an image of a clock https://files.catbox.moe/kgwa5e.png I asked it "what time does the clock show?" (both on reasoning: high) DS answered: The clock shows *5:10* (and 45 seconds). Here is the breakdown: * *Hour hand (red, shortest):* Pointing at the *5*. * *Minute hand (green, longest):* Pointing at the *2*, which represents 10 minutes.…

Gemini 3.7 Flash and 5.6-Sol (on all reasoning levels) also answer 8:10:25. The new "stealth" Ox Alpha also replies with the same. Opus 5 replies with 8:10 (no seconds). Not sure why this is so hard for them; Gemini is especially good at vision and I would have expected better from it.

Re: DeepSeek-v4-flash-vision-exp

#76

Post by nubg »

leumon wrote: post #74

It fails the simple clock test for me which Qwen3.8 27B got (nearly) right. given an image of a clock https://files.catbox.moe/kgwa5e.png I asked it "what time does the clock show?" (both on reasoning: high) DS answered: The clock shows *5:10* (and 45 seconds). Here is the breakdown: * *Hour hand (red, shortest):* Pointing at the *5*. * *Minute hand (green, longest):* Pointing at the *2*, which represents 10 minutes.…

welp, damning indictment. not sure if that means DS is super crap, or qwen is super good

Re: DeepSeek-v4-flash-vision-exp

#78

Post by andai »

knollimar wrote: post #72
johndough wrote: post #17

Earlier quoted context omitted.

Might still be fine. The most recent crop of vLLMs proactively use whichever programs are available on the system (e.g. ImageMagick or PIL) to "zoom in" by cropping subimages if they can't quite make out the details.

Downsizing a higher res image to lower res means the zoom will be blurry.

They process the original image file with Python on the local device. (And I've seen the web chats do this with their "computer use" features too.)

The really wild one is even blind models will do this and they'll try to run stats on the pixels to figure out what it looks like... the even wilder thing is that it kind of works!

Re: DeepSeek-v4-flash-vision-exp

#79

Post by ttul »

leumon wrote: post #74

It fails the simple clock test for me which Qwen3.8 27B got (nearly) right. given an image of a clock https://files.catbox.moe/kgwa5e.png I asked it "what time does the clock show?" (both on reasoning: high) DS answered: The clock shows *5:10* (and 45 seconds). Here is the breakdown: * *Hour hand (red, shortest):* Pointing at the *5*. * *Minute hand (green, longest):* Pointing at the *2*, which represents 10 minutes.…

A good share of humanity would have also gotten this question wrong!

Re: DeepSeek-v4-flash-vision-exp

#80

Post by jerkstate »

I just ran my image recognition benchmark on it ("is this XXX public landmark"?) and it misses a lot that bytedance seed 2.1 turbo gets right; for example: Asked "Is this Salisbury Cathedral" and supplied a picture of Wells Cathedral, it answers "Yes, the west facade of Salisbury Cathedral". Bytedance seed 2.1 turbo correctly says no. Similar results for a picture of Manhattan Bridge sent as Brooklyn Bridge, Chartres Cathedral sent as Notre Dame, etc. I have a benchmark of 12 such images and seed gets 11/12 and deepseek only gets 6/12.
Post reply on HN