Here are three news items that can speed up software delivery, cut costs, and expand AI capabilities in your applications.
DeepInfer released as open source
DeepSeek made the source code of DeepInfer publicly available, a high‑performance inference framework that powers its production services. The repo provides optimized kernels, model and sampler runtime, and attention‑DP parallelism.
Companies can adopt DeepInfer to run large language models on‑premises or in their cloud, lowering reliance on third‑party APIs and controlling inference costs. Consider migrating inference pipelines to DeepInfer or embedding it in new AI products.
How Dexpert helps: Backend & APIsSources
- GitHub (DeepSeek AI): DeepInfer: High-performance, large-scale inference framework for DeepSeek models
DeepSeek Search & Fetch API priced at $0.75 per 1k calls
DeepSeek introduced Search and Fetch APIs that give agents and RAG applications live web access, priced at $0.75 for every 1,000 successful requests.
The low price enables teams to add real‑time web lookup to generative AI flows without heavy spend, improving answer relevance. Try the API in assistant or recommendation system projects.
How Dexpert helps: Artificial IntelligenceSources
- DeepSeek: Search and Fetch API Documentation
Lambda Web Functions simplify HTTP API deployment
AWS added Lambda Web Functions, letting Lambda run a plain HTTP server without extra adapters or containers.
Developers can publish APIs faster, cutting provisioning time and operational costs. Evaluate moving existing HTTP services to Lambda Web Functions to streamline architecture.
How Dexpert helps: Cloud & ServerlessWant to put one of these to work in your company?
Dexpert builds custom systems, apps, integrations and AI solutions. Tell us what you need and get an estimate, or take the free digital maturity assessment.