Today in DeAI: DeepSeek retires V4-Flash in a silent swap and reverses its V4-Pro shutdown, NEAR AI Cloud ships TEE-attested inference, Chutes names a new CEO, and SaladCloud cuts consumer-GPU prices.
DeepSeek retires V4-Flash in a silent swap, spares V4-Pro
DeepSeek retired the legacy deepseek-v4-flash and deepseek-v4-flash-vision-exp IDs and now serves the newer V4.1-Flash behind them, and separately reversed a plan to route V4-Pro traffic to V4.1-Flash on September 14 after user pushback. Both changes appear only as footnotes on the pricing page. Why it matters: the model behind a pinned endpoint changed without a version bump — a trust signal every API builder should note. (DeepSeek pricing) — our coverage
NEAR AI Cloud ships OpenAI-compatible API with TEE attestations
NEAR AI Cloud now lets developers redirect OpenAI client code with a base-URL swap and routes to models hosted in Intel TDX and NVIDIA confidential-computing enclaves, with per-response attestation signatures. The privacy claims are NEAR's own, backed by a public verification path rather than an independent audit. Why it matters: it packages verifiable private inference as a standard API feature for open-weight models, not an enterprise add-on. (CryptoBriefing) — our coverage
Chutes names Christian Hyland CEO
Chutes, the Bittensor-based serverless inference network (SN64), appointed Christian Hyland — previously at Coinbase and Toku — as chief executive, per coverage this week. Why it matters: a senior hire signals the leading decentralized inference subnet is professionalizing as it scales, a step worth watching as the category matures. (Tao.media)
SaladCloud cuts consumer-GPU prices, raises top tier
SaladCloud cut prices 16–38% across many consumer GPUs effective September 12 — RTX 5080 High Priority dropped from $0.420 to $0.260/hr — while raising RTX 5090 High from $0.450 to $0.500/hr. Why it matters: cheaper consumer-GPU capacity shifts the self-hosting-versus-API cost math for bursty inference workloads. (SaladCloud)
Watching tomorrow
Whether DeepSeek publishes a formal migration notice for the V4-Flash retirement, and whether independent evaluators post V4.1-Flash results to test the launch-week benchmark claims.
Sources
- DeepSeek API Models & Pricing — DeepSeek
- NEAR AI launches OpenAI-compatible Cloud API with TEE attestations — CryptoBriefing
- Chutes names Christian Hyland CEO as AI compute platform expands — Tao.media
- SaladCloud price changes, September 2026 — SaladCloud
About DeAI
DeAI is an independent publication covering open-weight AI models, private inference, and decentralized infrastructure — the tools for running AI you actually control. We test providers on price, privacy, and refusal behavior and publish the numbers, not the vibes. DeAI is powered by Morpheus (mor.org), a decentralized inference marketplace, and covers it on the same terms as every other provider.
Powered by Morpheus and StrandCMS
Morpheus is a decentralized inference marketplace, covered on the same terms as every other provider — we rank it wherever the data lands. StrandCMS is the open-source, agent-first framework this site is built on.