DeepSeek V4 Pro 0813
225 points - today at 4:04 PM
SourceComments
eshack94 today at 5:40 PM
It appears that the only available endpoint (as of this writing) requires enabling "Allow paid endpoints that train on request data" in the OpenRouter privacy settings. I hope additional paid providers will become available that don't require training on data.
scrlk today at 4:42 PM
Benchmarks:
| Benchmark | DS-V4-Pro | DS-V4-Flash | DS-V4-Pro | DS-V4-Flash | GLM-5.2 | Kimi-K3 | Opus-4.8 | Fable 5 |
| | 0813 | 0731 | Preview | Preview | | | | (w/ fallback) |
|--------------------------|-----------|-------------|-----------|-------------|-----------|-----------|-----------|---------------|
| HLE (wo/w tools) | 42.7/60.0 | 37.8/51.5 | 37.7/48.2 | 34.8/45.1 | 40.5/54.7 | 43.5/56.0 | 49.8/57.9 | 53.3/63.0 |
| Terminal Bench 2.1 | 87.9 | 82.7 | 72.1 | 61.8 | 81.0 | 88.3 | 85.0 | 88.0 |
| NL2Repo | 61.5 | 54.2 | 38.5 | 39.4 | 48.9 | - | 69.7 | - |
| Cybergym | 83.3 | 76.7 | 52.7 | 38.7 | - | 80.0 | 78.3 | 83.1 |
| DeepSWE | 62.7 | 54.4 | 12.8 | 7.3 | 46.2 | 67.5 | 58.0 | 70.0 |
| Toolathlon-Verified | 74.1 | 70.3 | 55.9 | 49.7 | 59.9 | 76.5 | 76.2 | 77.9 |
| Agents' Last Exam | 25.7 | 25.2 | 16.5 | 15.8 | 23.8 | 27.6 | 25.7 | - |
| AutomationBench (Public) | 31.8 | 25.1 | 12.8 | 10.8 | 12.9 | 30.8 | 27.2 | 29.1 |
| DSBench-FullStack | 71.1 | 68.7 | 41.8 | 37.0 | 61.8 | 73.7 | 71.6 | 77.2 |
| DSBench-Hard | 67.2 | 59.6 | 31.1 | 25.8 | 54.5 | 63.0 | 71.7 | 68.3 |
alecsm today at 5:24 PM
I've been using the last Deepseek Flash update for a week and I'm amazed. It was a capable model for easy tasks but now it looks like it can do some heavy development for peanuts.
I can't wait to try this new one.
jklmnopqrstuvw today at 5:34 PM
Tested both DS v4 pro 0813 and Grok 4.6 (all from openrouter) on Codex cli. Worked on a same new feature development on my project.
Deepseek 4 pro: Worked for 12m 02s - cost $0.12 - has bug.
Grok 4.6: Worked for 3m 18s - cost $ 1.41 - no bug.
BTW, why grok 4.6 news being down voted and disappeared from frontpage?
aabdi today at 4:06 PM
https://api-docs.deepseek.com/quick_start/pricing/
Competitive with opus 4.8 but weaker than sol or fable. About 20x cheaper.
Gecko4072 today at 4:48 PM
Currently burning money quickly on official deepseek api. They are also increasing pricing starting today. V4 Flash 0731 still feels like the most outstanding model of the past few months and probably to come.
indigodaddy today at 4:38 PM
@dang - Pls merge this with https://news.ycombinator.com/item?id=49274018
book_mike today at 5:12 PM
What I care about is whether the model is capable of the tasks I give it at the lowest cost. Right now I'm using Kimi-K3/GLM-5.2/Minimax. Sonnet is great but I burn through the tokens too fast. Opus 5 set to max is amazing and more intelligent than all of us. .998 of the time I don't need that kind of intelligence. I just need the job done.
nthypes today at 5:32 PM
Still behind Kimi-K3 in almost half of the benchmarks
Readerium today at 5:23 PM
V4 Pro has vision correct?
deleted today at 4:19 PM
LeonKnst today at 4:41 PM
I find it interesting how much adoption seems to be influenced by momentum. Some of these Chinese models are surprisingly capable, but developers often default to the models that are already established as the “industry standard
yipinwong today at 5:31 PM
Worse than Luna but more expensive than Luna.
Sticking with Luna without sending my data to Deepseek (China)