DeepSeek-V4's Ascend pivot: cheaper tokens, shakier answers

Sources

DeepSeek-V4: a million-token context that agents can actually use huggingface.co

References

smol.ai / Artificial Analysis writeup news.smol.ai

DeepSeek-V4-Pro recorded a 94% hallucination rate (V4-Flash 96%) on AA-Omniscience because it almost never abstains when uncertain, in contrast to GLM-5 which lowered hallucinations via a refusal-to-answer mechanism.

Skywork.ai independent eval skywork.ai

Failures in retrieval and reasoning appear randomly across different positions in the 1M window rather than clustering in a fixed weak zone… attributed to the Lightning Indexer and sparse attention sporadically missing specifics during compression.

Lushbinary analysis lushbinary.com

V4-Pro and V4-Flash were trained and deployed exclusively on Huawei Ascend 950PR silicon, completely abandoning the Nvidia CUDA ecosystem used for previous versions.

Fireworks.ai engineering blog fireworks.ai

DSec exposes four execution substrates—function calls, containers, microVMs (Firecracker), and full VMs (QEMU)—behind a unified Python SDK, backed by 3FS at 6.6 TiB/s peak read and 40 GiB/s for KV cache I/O, with token-granular write-ahead logs enabling preemption-safe trajectory replay.

Medium review (leucopsis) medium.com

The reasoning_content generated in thinking mode must be passed back to the API in all subsequent turns when tool calls are active… many existing tools built for OpenAI-style completions don’t expect a separate reasoning field, creating an integration hurdle in IDEs like Cursor.

Artgor Medium review artgor.medium.com

While V4 rivals frontier models in coding and tool use, it still trails the top closed-source models like GPT-5.5 and Gemini 3.1 Pro by three to six months in general reasoning, with a visible gap on ARC-AGI where V4-Pro remains meaningfully behind the US frontier.