ai/intel

Read

  • Today
  • Archive
  • Briefs
  • Labs

You

  • Settings

Live · updated 6:27 AM

ai/intel
Labs / DeepSeek
D

DeepSeek

35 stories in the last 30 days

Models in the top ranksLMArena text, 2026-11-19

#42DeepSeek-V4.1-Flash-20260910 (Max)1,475
#65DeepSeek-V4-Pro (High, 20260813)1,464
#72DeepSeek-V4-Pro1,458
#81DeepSeek-V4-Pro Preview (High)1,455

Stories

Thu 19 Nov

4:09 AM

India Startup Sells DeepSeek Access While Pursuing Sovereign AI

DeepSeek · Pricing

Sat 14 Nov

7:12 PM

Update: tandem.cpp — DeepSeek-V4.1 on a 64 GB PC with a 16 GB card, and AMD (ROCm) support

DeepSeek · Models
3:01 PM

DeepSeek API Adds Search, OCR and Speech: 120-Second Audio Cap, Vendor Bound by Chinese Laws

DeepSeek · Policy
12:50 PM

Lawyer’s eight-V100 AI server moves DeepSeek from 2-bit to near-NVFP4 precision

Fri 13 Nov

5:39 PM

DeepSeek Ultrafast

DeepSeek · Models
4:31 PM

DeepSeek flash max is the first model to reach >1% on challenging SWE-sweep bug finding benchmark

DeepSeek · Models

Thu 12 Nov

11:31 PM

Lawyer with too many GPUs, part 5: DeepSeek-V4.1-Flash at near-NVFP4 quality on eight V100s from 2018. 524K context, 100+ tok/s decode, ~160 tok/s with two chats at once, and my legal harness now runs on it, 100% local

DeepSeek · Chips
12:50 AM

DeepSeek V4.1 Flash 1017 exl3 2.9bpw APEX 2025 benchmark results

DeepSeek · Models

Wed 11 Nov

4:47 PM

DeepSeek has Two Numbers for Active Parameter Count. How?

DeepSeek · Models
4:17 PM

New Qwen3.8-27B Unsloth Dynamic V3.1 GGUFs

DeepSeek · Open weights

Tue 10 Nov

11:50 PM

XeMaya, an Intel Arc inference engine to run DeepSeek V4.1 and GLM 5.3

DeepSeek · Models

Mon 9 Nov

4:27 PM

RustingStrata vs llama.cpp, DeepSeek-V4.1-Flash Q2 (249 GB) on a single RTX 3060 12 GB with 30 GB RAM

DeepSeek · Open weights

Sun 8 Nov

6:01 PM

DeepSeek’s AI Agent Platform Hits 1.54 Million Daily Users

DeepSeek · Models

Sat 7 Nov

7:58 PM

Betelgeuse running DeepSeek V4.1 Flash Q8_XL (1017) at 19 tok/s decode & 62 tok/s prefill on an RTX 3060 12GB + 16GB RAM: How we finally fixed the prefill bottleneck on out-of-core MoEs

DeepSeek · Models

Fri 6 Nov

6:43 PM

DeepSeek Nearly Matches OpenAI Cyber Model in Simulated Network Tests

DeepSeek · Open weights

Thu 5 Nov

9:07 PM

Lawyer with too many GPUs, part 4: DeepSeek-V4.1-Flash on eight V100s from 2018. 200+ tok/s on code, ~110 on legal writing, 2,200 tok/s prefill at 128K, 524K context, every weight in VRAM

DeepSeek · Chips
6:40 PM

Deepseek fixed my HDMI display show case

DeepSeek · Models
6:23 PM

Delta Inference Engine v0.93 adds Unsloth quant support and lower-RAM operation

Wed 4 Nov

8:39 PM

Deepseek accounted for 53% of tokens on OpenRouter, community post says

DeepSeek · Open weights

Mon 2 Nov

3:57 PM

Unsloth reports DeepSeek-V4.1 GGUF benchmarks above GSQ-RCO on several tests

DeepSeek · Open weights
3:52 PM

Atomic.Chat team releases refusal-edited DeepSeek model for Strata

DeepSeek · Models
2:06 PM

DeepSeek plans to train its V5 model on Huawei Ascend 950 chips

Sat 31 Oct

4:25 PM

DeepSeek releases the open source prototype of its Super-agent

DeepSeek · Open weights
5:32 AM

Lawyer with too many GPUs, part 3: DeepSeek-V4.1-Flash on 2x DGX Spark + an old 3090 box. 2,600+ tok/s prefill to 256K, 322 tok/s structured output, full stack and raw data published

DeepSeek · Chips
1:09 AM

QEngine 1.3.1 leads tested engines in Qwen3.8-Flash-Next decode on two RTX 5090s

Fri 30 Oct

4:46 PM

ISTA attributes DeepSeek benchmark gap to different reasoning-output limits

DeepSeek · Models
1:48 PM

Introducing DeepSeek-V4.1-Flash-1017 Dynamic v3.1 Unsloth GGUFs

DeepSeek · Models
1:12 PM

Tried to get Qwen 3.8 27b to make the deepseek chibi whale in blender and animate it lol

Thu 29 Oct

5:46 PM

Why you can't trust cloud API endpoints: 40 days of coding-agent telemetry showing the exact day providers silently overwrote deepseek-v4.1-flash with -1017 ("faster execution, shorter answers" = 3x more swearing and 6.7x more model switches)

DeepSeek · Models

Wed 28 Oct

12:39 AM

DeepSeek-V4.1-Flash-1017 at 41 tok/s on one 3090

DeepSeek · Models

Tue 27 Oct

10:51 AM

DeepSeek AI model shows safety gaps in independent tests

DeepSeek · Models
4:53 AM

There are now 5 inference providers for abliterated DeepSeek

DeepSeek · Models
Filter these in the archive →
DeepSeek · Models
DeepSeek · Chips
DeepSeek · Chips
1:54 PM

DeepSeek Adds Plugin-Creation Entry to Harness Creator Mode

DeepSeek · Business
1:10 PM

User praises DeepSeek V4.1 Flash 1017 for speed and reliability

DeepSeek · Open weights
12:44 AM

Developer patches Strata to reduce DeepSeek V4.1 agent prompt-processing delays

DeepSeek · Chips
DeepSeek · Models
DeepSeek · Models