--- title: "#deepseek articles | Hashnode" description: "Explore 345 articles tagged #deepseek on Hashnode. Tutorials, guides, and how-tos from the community." url: "https://hashnode.com/tag/deepseek" --- Tag # #deepseek 345 posts · 25 followers [ModelSite](https://hashnode.com/@modelsite) · [2d ago](https://hashnode.com/posts/call-deepseek-qwen-and-glm-with-one-openai-compatible-api-key/6af28cf05fe0e64105d62624) ### [Call DeepSeek, Qwen and GLM with one OpenAI-compatible API key](https://modelsite.hashnode.dev/call-deepseek-qwen-and-glm-with-one-openai-compatible-api-key?utm_source=hashnode&utm_medium=feed) Same OpenAI SDK, one endpoint: change the model name and keep your code. 2 min read · [modelsite.hashnode.dev](https://modelsite.hashnode.dev/call-deepseek-qwen-and-glm-with-one-openai-compatible-api-key?utm_source=hashnode&utm_medium=feed) [Kris Holy](https://hashnode.com/@kris-holy) · [2d ago](https://hashnode.com/posts/ai-inference-platforms-benchmark-2026/6af23714bb2d9123f95fba2f) ### [AI Inference Platforms Benchmark 2026: Fireworks, Together, Baseten, Modal, DeepInfra, Z.ai and Keln](https://keln.hashnode.dev/ai-inference-platforms-benchmark-2026?utm_source=hashnode&utm_medium=feed) We replayed real agent sessions on DeepSeek V4.1 Flash and GLM-5.3 Flash and compared speed, reliability, prompt caching and cost. 10 min read · [keln.hashnode.dev](https://keln.hashnode.dev/ai-inference-platforms-benchmark-2026?utm_source=hashnode&utm_medium=feed) [xiun xiao](https://hashnode.com/@xiunxiao) · [2d ago](https://hashnode.com/posts/deepseek-v4-1-flash-vs-claude-haiku-5-5-a-17-task-benchmark-of-the-two-cheapest-tiers/6af273133f36efab05c00df0) ### [DeepSeek V4.1 Flash vs Claude Haiku 5.5: a 17-task benchmark of the two cheapest tiers](https://xiunxiao.hashnode.dev/deepseek-v4-1-flash-vs-claude-haiku-5-5-a-17-task-benchmark-of-the-two-cheapest-tiers?utm_source=hashnode&utm_medium=feed) The two cheapest tiers on the same 17 coding, extraction and reasoning tasks, three runs each: pass rate, latency and cost per 1,000 tasks. 6 min read · [xiunxiao.hashnode.dev](https://xiunxiao.hashnode.dev/deepseek-v4-1-flash-vs-claude-haiku-5-5-a-17-task-benchmark-of-the-two-cheapest-tiers?utm_source=hashnode&utm_medium=feed) [KarmaBox](https://hashnode.com/@karmabox) · [2d ago](https://hashnode.com/posts/a-284b-model-on-a-64gb-pc-has-to-beat-0-00086-a-task-here-s-the-math/6af1ad53d68d397ce466df28) ### [A 284B Model on a 64GB PC Has to Beat $0.00086 a Task. Here's the Math](https://karmabox.hashnode.dev/a-284b-model-on-a-64gb-pc-has-to-beat-0-00086-a-task-here-s-the-math?utm_source=hashnode&utm_medium=feed) llama-modchip fits DeepSeek-V4-Flash into 64GB of RAM with no GPU. We priced that box against per-token billing, with every assumption shown. 20 min read · [karmabox.hashnode.dev](https://karmabox.hashnode.dev/a-284b-model-on-a-64gb-pc-has-to-beat-0-00086-a-task-here-s-the-math?utm_source=hashnode&utm_medium=feed) [李弘基](https://hashnode.com/@kd-agentic) · [5d ago](https://hashnode.com/posts/deepseek-superagent-don-t-chat-in-one-window-hand-work-to-a-repo-worker/6aee098b9c8705fdf758067f) ### [DeepSeek Superagent: Don't Chat in One Window. Hand Work to a Repo Worker.](https://kd-agentic.hashnode.dev/deepseek-superagent-don-t-chat-in-one-window-hand-work-to-a-repo-worker?utm_source=hashnode&utm_medium=feed) Check the environment, send one small GitHub task, and review the result before expanding its responsibilities. 13 min read · [kd-agentic.hashnode.dev](https://kd-agentic.hashnode.dev/deepseek-superagent-don-t-chat-in-one-window-hand-work-to-a-repo-worker?utm_source=hashnode&utm_medium=feed) [Jakson Tate](https://hashnode.com/@jakson-tate) · [5d ago](https://hashnode.com/posts/deepseek-deepinfer-bare-metal-deployment/6aed825df20a479c8e9cc367) ### [DeepSeek DeepInfer: Production Inference Architecture, DSpark Optimization & Bare-Metal GPU Deployment](https://servermo.hashnode.dev/deepseek-deepinfer-bare-metal-deployment?utm_source=hashnode&utm_medium=feed) An SRE guide to DeepSeek-V4.1-Flash inference, SWA bounded replay, speculative decoding, and production serving requirements. 6 min read · [servermo.hashnode.dev](https://servermo.hashnode.dev/deepseek-deepinfer-bare-metal-deployment?utm_source=hashnode&utm_medium=feed) [ming](https://hashnode.com/@pipipi) · [Nov 3](https://hashnode.com/posts/does-it-save-tokens-tokeneconomizer-on-deepseek-harness/6ae9e38f326dcadfe3937a3b) ### [Does It Save Tokens? TokenEconomizer on DeepSeek Harness](https://bigai.hashnode.dev/does-it-save-tokens-tokeneconomizer-on-deepseek-harness?utm_source=hashnode&utm_medium=feed) I ported TokenEconomizer to DeepSeek Harness and ran 12 A/B sessions. Total tokens: 209,263 with the plugin, 208,239 without, a 0.5% gap. 8 min read · [bigai.hashnode.dev](https://bigai.hashnode.dev/does-it-save-tokens-tokeneconomizer-on-deepseek-harness?utm_source=hashnode&utm_medium=feed) [99 sunny](https://hashnode.com/@sunny99x) · [Nov 1](https://hashnode.com/posts/native-multimodal-does-not-mean-one-backbone/6ae77aeace2a769e427fb89c) ### [Native multimodal still allows a separate encoder](https://sunny99.hashnode.dev/native-multimodal-does-not-mean-one-backbone?utm_source=hashnode&utm_medium=feed) Native multimodal does not mean dropping separate vision encoders; DeepSeek-V4.1-Flash still uses ViT + projector, and the real goal is an updatable world state. 7 min read · [sunny99.hashnode.dev](https://sunny99.hashnode.dev/native-multimodal-does-not-mean-one-backbone?utm_source=hashnode&utm_medium=feed) [Magnus](https://hashnode.com/@magnus98) · [Nov 1](https://hashnode.com/posts/an-open-source-desktop-agent-that-follows-deepseek-s-super-agent-design-and-what-we-left-out/6ae7699873cabd3ede128433) ### [An open-source desktop agent that follows DeepSeek's super-agent design (and what we left out)](https://modelflare.hashnode.dev/an-open-source-desktop-agent-that-follows-deepseek-s-super-agent-design-and-what-we-left-out?utm_source=hashnode&utm_medium=feed) Disclosure: I maintain AnyModelBot. This is an unofficial project, not affiliated with DeepSeek. Most chatbots wait for you to speak and stop when they answer. Plenty of real work is not a question an 6 min read · [modelflare.hashnode.dev](https://modelflare.hashnode.dev/an-open-source-desktop-agent-that-follows-deepseek-s-super-agent-design-and-what-we-left-out?utm_source=hashnode&utm_medium=feed) [CloudGPU](https://hashnode.com/@cloudgpu) · [Oct 29](https://hashnode.com/posts/roo-code-profiles-deepseek-kimi-minimax/6ae3178af2c7e40801cd2cec) ### [Roo Code configuration profiles for DeepSeek, Kimi and MiniMax on one OpenAI-compatible endpoint](https://cloudgpu.hashnode.dev/roo-code-profiles-deepseek-kimi-minimax?utm_source=hashnode&utm_medium=feed) One base URL and one key, a profile per model ID, a profile per mode, and a tool-call check for each 6 min read · [cloudgpu.hashnode.dev](https://cloudgpu.hashnode.dev/roo-code-profiles-deepseek-kimi-minimax?utm_source=hashnode&utm_medium=feed)