# DeepSeek Makes Inference Economics the New Frontier

*By AI High Signal Digest • September 11, 2026*

DeepSeek’s V4.1-Flash turns architecture-level efficiency into a direct challenge on price-performance, while an Anthropic threat report and a new RSI benchmark sharpen the operational and safety picture.

## Top Stories

*Why it matters: Frontier advantage is increasingly expressed as cost, throughput and control of real-world use—not parameter count alone.* [^1][^2]

**DeepSeek’s V4.1-Flash makes inference economics the headline.** DeepSeek describes a 552B MoE with a Causal Encoder–Decoder activating 8B parameters for input and 16B for output, while its KV cache uses one-quarter the prior HBM and one-eighth the SSD. [^3][^4] Artificial Analysis gives it a 40 Intelligence Index score, says it beats the 1.6T V4-Pro at roughly four times lower per-token cost, and estimates $0.27 per task despite 89k tokens per task. [^1] DeepSeek will route V4-Pro requests to Flash at Flash rates from September 14 until V4.1-Pro launches. [^5]

**Anthropic’s threat report makes misuse operational.** It covers attempted use of Claude for cyberattacks, influence operations, surveillance, biology and weapons; Anthropic says it disrupted every operation described, strengthened safeguards and shared findings with authorities and other AI companies. The cases are atypical, but the company calls them among its most sophisticated examples of where AI misuse is heading. [^6]

**The RSI Index tempers recursive-self-improvement claims.** ValsAI, Marimo and CoreWeave say their first third-party benchmark finds that frontier models can perform AI-research tasks but remain far from the human frontier. [^2] No agent reached the reference result on any task; Fable 5.1 led at 35%, while reproducing one language-model recipe took 30 minutes on one H100 versus roughly 175 hours for the published TPU result. [^7] The systems mostly recombined known techniques, and the initial results are single fixed-budget runs. [^8][^9]

## Research & Innovation

*Why it matters: Agent training is becoming a control problem—how long to interact, and how to preserve model–harness fit.*

**Qwen’s Elastic Horizon** uses the 90th percentile of successful trajectory lengths to detect when extra environment interactions stop improving outcomes. The paper reports the best success rates across 7B and 14B backbones and up to 25% fewer per-step trajectory tokens. [^10]

**Salesforce’s co-evolution study** found that fine-tuning a weak model on expert trajectories after its harness had evolved reduced performance by 4–30 points across seven enterprise tasks. Its proposed fix rewrites only the failing turn, preserving the weaker model’s planning style instead of copying a full expert rollout. [^11]

## Products & Launches

*Why it matters: Agent vendors are packaging persistent state, delegation and execution environments—not just chat endpoints.*

**GPT-Live-1** is available in the API for voice agents that listen while speaking; OpenAI describes controllable delegation and a price of $0.05 per minute. Paired with Astra, it completed 83.6% of customer-support tasks on the first attempt, versus 45.7% for Realtime 2.1. [^12][^13][^14]

**OpenAI’s Agents API** is in public beta: OpenAI manages Codex orchestration, long-running sessions and context, while developers choose the agent’s capabilities and execution environment. OpenAI-hosted sandboxes can run code, work with files and produce artifacts. [^15][^16][^17]

**Cursor Projects** puts a coordinator agent in a persistent thread; it can schedule tasks, follow pull requests, monitor Slack, and share memory and artifacts across devices. The feature is rolling out in beta. [^18][^19][^20][^21]

## Industry Moves

*Why it matters: The race is now to own the stack that lets agents run continuously and cheaply.*

**Baseten and Blaxel are combining model infrastructure with agent execution.** Blaxel is joining Baseten to add isolated microVM sandboxes, persistent storage and production networking to model serving and training; its sandboxes suspend and resume in 25 milliseconds, which the company claims is up to five times faster than alternatives. [^22] Their stated end state is one system for agent execution, inference and training. [^22]

**OpenAI Foundation** committed $60 million over three years to bring AI weather and crop-disease forecasts to 100 million smallholder farmers across South and Southeast Asia and East Africa, working with governments and local institutions. [^23]

**Positron** said it raised $875 million at a $5 billion valuation for AI-acceleration hardware. [^24]

## Quick Takes

*Why it matters: Capability and price-performance gains are spreading across models, benchmarks and specialized systems.*

- **Math:** Epoch AI says GPT-6 Astra solved the last FrontierMath Tier 4 problem; the benchmark rose from 5% to 98% in under 14 months and is now considered saturated. [^25][^26]
- **Open models:** Tencent Hunyuan’s Hy4 preview ranked second among open models across 14.5K+ real-world agent sessions, at a median $0.26 per task versus $0.80 for Kimi K3 Max. [^27]
- **Coding:** Cognition says SWE-2 reached 50% on FrontierCode, matching Fable 5.1 at 64% lower cost; it is free in Devin for Pro, Max and Teams subscribers for one month. [^28][^29]

---

### Sources

[^1]: [𝕏 post by @ArtificialAnlys](https://x.com/ArtificialAnlys/status/2098148674203488422)
[^2]: [𝕏 post by @ValsAI](https://x.com/ValsAI/status/2098170083466191086)
[^3]: [𝕏 post by @deepseek_ai](https://x.com/deepseek_ai/status/2097930613101838709)
[^4]: [𝕏 post by @deepseek_ai](https://x.com/deepseek_ai/status/2097930617396887773)
[^5]: [𝕏 post by @deepseek_ai](https://x.com/deepseek_ai/status/2097930620680941732)
[^6]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2098097512544444447)
[^7]: [𝕏 post by @ValsAI](https://x.com/ValsAI/status/2098170085273924003)
[^8]: [𝕏 post by @ValsAI](https://x.com/ValsAI/status/2098170087102681243)
[^9]: [𝕏 post by @ValsAI](https://x.com/ValsAI/status/2098170090118377522)
[^10]: [𝕏 post by @dair_ai](https://x.com/dair_ai/status/2098109386568925397)
[^11]: [𝕏 post by @omarsar0](https://x.com/omarsar0/status/2097958286146605446)
[^12]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2098099269551149398)
[^13]: [𝕏 post by @juberti](https://x.com/juberti/status/2098102270332444922)
[^14]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2098118255449977157)
[^15]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2098130570048045453)
[^16]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2098130589136425339)
[^17]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2098130625144451444)
[^18]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2098162488013455784)
[^19]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2098162490026697134)
[^20]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2098162491754741780)
[^21]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2098162493663146411)
[^22]: [𝕏 article by @baseten](https://x.com/i/article/2098081306425090048)
[^23]: [𝕏 post by @FoundationOAI](https://x.com/FoundationOAI/status/2098065380518347260)
[^24]: [𝕏 post by @WSJTech](https://x.com/WSJTech/status/2098034387921010967)
[^25]: [𝕏 post by @EpochAIResearch](https://x.com/EpochAIResearch/status/2098103831502708864)
[^26]: [𝕏 post by @EpochAIResearch](https://x.com/EpochAIResearch/status/2098103846191186203)
[^27]: [𝕏 post by @arena](https://x.com/arena/status/2098116858180509843)
[^28]: [𝕏 post by @cognition](https://x.com/cognition/status/2098069238044860487)
[^29]: [𝕏 post by @cognition](https://x.com/cognition/status/2098069247842812103)