DeepSeek Flash inverted the economics of agent products
There is an adversarial relationship between developers and big model labs. Model labs charged developers higher API prices to subsidize their own agent harness offerings. Think Anthropic charging 5x higher Claude API prices to subsidize consumer subscriptions. So Cursor in a way was subsidizing their own direct competitor. DeepSeek V4 Flash totally inverted this relationship. Now you have a model that beats even Sonnet in some benchmarks and is totally opensourced. Now inference providers are racing to the bottom to optimize and give cheaper hosting. Every player with a non-SOTA is now…
What it does
In the maker’s words, at launch
There is an adversarial relationship between developers and big model labs. Model labs charged developers higher API prices to subsidize their own agent harness offerings. Think Anthropic charging 5x higher Claude API prices to subsidize consumer subscriptions. So Cursor in a way was subsidizing their own direct competitor. DeepSeek V4 Flash totally inverted this relationship. Now you have a model that beats even Sonnet in some benchmarks and is totally opensourced. Now inference providers are racing to the bottom to optimize and give cheaper hosting. Every player with a non-SOTA is now racing to swap over to stop paying the big model lab tax, even Microsoft is switching Copilot to use DeepSeek. On switching over to Deepseek: - we noticed over a 100x cost decrease while similar or better performance then Gemini 3 Flash - insane saving from the cached input tokens: $0.002/1 Million tokens - both DeepSeek Flash and GLM 5.2 are text-only models, so clearly multimodal training is not worth the additional cost. Language is just a much more efficient sparse representation of the world/reasoning than vision - we had a early bet on a text-only web agent harness, and now with DeepSeek this results in unique cost advantages. - we rewrote our harness as a callable DSL library that a model can generate code to execute on. DeepSeek has proven phenomenal on code generation to drive an agent harness. - I would highly recommend everyone to rewrite their harness to be text-only and callable via executable code leveraging DeepSeek V4 Flash.
Does the same job
all alternatives →
- DDDistilling DeepSeek into GPT-OSS doesn't transfer censorship. Try itJul 2026 · ctgt.ai · ▲170
We recently used DeepSeek V4 Flash as a teacher for finance tasks with GPT-OSS-120B. Distillation works well on this problem. At a constrained 8k token budget, our self-distilled 120B scores 83.61% on FinanceReasoning, above Kimi K3 (81.93%) and Inkling (65.13%). We released the 20B open weights. With V4 as the teacher though, we realized it would be timely to measure if the censorship characteristic of it transferred to the distilled version of the base model. tl;dr it didn't, the teacher answered politically sensitive questions 7 SDs differently than expected, but the distilled model's…
- SSsllm – Split a GPU node with other developers, unlimited tokensApr 2026 · sllm.cloud · ▲188
Running DeepSeek V3 (685B) requires 8×H100 GPUs which is about $14k/month. Most developers only need 15-25 tok/s. sllm lets you join a cohort of developers sharing a dedicated node. You reserve a spot with your card, and nobody is charged until the cohort fills. Prices start at $5/mo for smaller models. The LLMs are completely private (we don't log any traffic). The API is OpenAI-compatible (we run vLLM), so you just swap the base URL. Currently offering a few models.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, June 2026
the whole month →
Fundraisly▲1,544AI fundraising agent that finds investors and books meetings
AI · Jun 2026 · fundraisly.com
- H6Homebrew 6.0.0▲1,481
Today, I’m proud to announce Homebrew 6.0.0. The most significant changes since 5.1.0 are a new tap trust security mechanism, the new faster, smaller, default internal Homebrew JSON API, sandboxing on Linux, better defaults informed by our user survey, many brew bundle improvements, improved performance and initial support for macOS 27 (Golden Gate). Happy to discuss any questions here!
Dev tools · Jun 2026 · brew.sh
- PU
hope you enjoy
Life & fun · Jun 2026 · vorpus.github.io
- IM
Life & fun · Jun 2026 · hackernewstrends.com