Live

Companies & models

DeepSeek

Model releases, open-source weights, technical reports, and API pricing changes of DeepSeek (Deep Exploration).

7featured stories

6this month

14 Nov, 21:05Updated

Active this month

Major events

14 Nov
DeepSeek has opened the entire suite of application APIs, with voice interfaces included for the first time. 2 sources
DeepSeek has officially opened the entire suite of application APIs, including online search, deep thinking search, file upload, and link reading features within the app, all providing independent interfaces that can be called separately and billed individually. Voice capabilities have been included in the open list for the first time, including ASR and TTS interfaces. The original price for speech recognition is 0.78 yuan per hour, now half off at 0.39 yuan per hour, and speech synthesis is 1.5 yuan per 10,000 characters.
12 Nov
Doubao Work has integrated models such as Kimi 3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash, and is offering limited-time promotions. 2 sources
Dou Bao has officially integrated multiple third-party models such as Kimi 3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash, and has launched a limited-time special offer plan. Users can select the corresponding model according to their actual work needs to complete tasks.

Latest stories

Nov 14Sat1–7 / 7

Nov 14Saturday1–7 / 7

21:05
QbitAI· 21:05Featured8080
DeepSeek has opened the entire suite of application APIs, with voice interfaces included for the first time.
DeepSeek has officially opened the entire suite of application APIs, including online search, deep thinking search, file upload, and link reading features within the app, all providing independent interfaces that can be called separately and billed individually. Voice capabilities have been included in the open list for the first time, including ASR and TTS interfaces. The original price for speech recognition is 0.78 yuan per hour, now half off at 0.39 yuan per hour, and speech synthesis is 1.5 yuan per 10,000 characters.
Products #Product updates #DeepSeek #Voice
Why it matters: The original text lists the tool interfaces opened by DeepSeek and the half-price voice pricing, allowing readers to estimate the call costs for voice assistant and document processing applications.

Nov 13Fri

Nov 13Friday

11:32
Hacker News · AI hot posts· 11:32Featured8080
Unemployment Labs reports that the models DeepSeek, Kimi, MiniMax, and Ant have weights with a Blackwell hardware fingerprint.
Unemployment Labs, by analyzing the random number initialization layout in public model weights, discovered DeepSeek-V4-Flash, DeepSeek-V4-Pro, Kimi K3.
Research #Papers & research #Safety & alignment #Data & training
Why it matters: By reversing the random number layout in the model weights to train the hardware, it provides a reproducible technical evidence path for export control disputes.

Nov 9Mon

Nov 9Monday

05:40
Hacker News · 300+· 05:40Featured7878
llama-modchip enables the 284B DeepSeek-V4-Flash to run on a 64GB memory without GPU
llama-modchip releases a set of ggml/llama.cpp core patches and quantization transcoding tools, compressing the 284.33B DeepSeek-V4-Flash to 56.56 GiB (1.71 bpw), which can run on a pure CPU machine with 64 GB memory, with a peak memory of 57.61 GiB under 256K context and no need for swap.
Guides #Tutorials #Engineering #Reasoning
Why it matters: The project provides a quantization scheme and core patches for running 284B MoE on 64 GB memory, along with measured data for each hardware, which can be used to determine the feasible boundary for pure CPU deployment.

Nov 6Fri

Nov 6Friday

23:40
vLLM Blog· 23:40Featured7272
REINFORCE Zero: Aligned kernel allows asynchronous RL to train stably under lag 64.
NVIDIA Research and Inferact's REINFORCE Zero proposes to first align rollout with training kernel, then use unclipped token-level importance sampling to correct weight lag, achieving a training step throughput 2.40 times that of synchronous training under 16 GPUs and 12K updates.
Research #Papers & research #Reasoning #Data & training
Why it matters: The paper breaks down the gap between asynchronous RL's rollout and training into two items: weight lag and kernel inconsistency, and proposes an alignment-free, no-trimming importance sampling method after alignment.

Nov 1Sun

Nov 1Sunday

07:35
QbitAI· 07:35Featured7878
DeepSeek has open-sourced the Super-agent architecture, disclosing the underlying concept of Personal Agent.
DeepSeek has open-sourced the Super-agent prototype, disclosing the dual architecture concept for Personal Agent, where the upper-layer Router handles message reception and task coordination, and the lower-layer Worker executes tasks in parallel.
Research #Papers & research #Agent #Open source
Why it matters: DeepSeek has publicly disclosed the dual architecture of Personal Agent and actual test data, showing how multi-session collaboration can bypass human attention bottlenecks.

Oct 25Sun

Oct 25Sunday

02:42
vLLM Blog· 02:42Featured6060
vLLM serves DeepSeek-V4.1-Flash on GB300 NVL72: performance matches DeepInfer and extends to the entire rack
The vLLM team used P/D separation and Wide-EP to serve DeepSeek-V4.1-Flash on a single NVIDIA GB300 NVL72 rack, achieving 954K tokens/s per GPU and 196 tokens/s for single-user P90 on SemiAnalysis AgentX, which is 7.9 times the early single-node results.
Guides #Tutorials #Engineering #Reasoning
Why it matters: The vLLM team has released per-item optimization and measured data for serving DeepSeek-V4.1-Flash on GB300 NVL72, which can be compared with DeepSeek's own engine DeepInfer.

Oct 7Wed

Oct 7Wednesday

00:00
vLLM engineering· 00:00Featured4242
vLLM explains a 5.3× throughput gain for DeepSeek-V4.1-Flash
Inferact and the vLLM team report 1.9× faster low-concurrency serving and 5.3× higher throughput under a 150-tokens-per-second constraint after three weeks of optimization. The changes combine bounded replay of sliding-window attention, CUDA graphs, and fused kernels. Their post explains how cache reconstruction and skipping parts of prompt prefill reduce serving work.
Guides
Why it matters: Editorial takeaway: model benchmarks alone do not describe serving costs. This engineering breakdown helps identify cache and kernel bottlenecks; the reported gains apply to its benchmark conditions, rather than every deployment.