Price cuts sweep AI models as n8n and LoopX ship fixes
28,702 stars today66 new projects+3,112 dsh-Go
By olud.ai editorial · built from our own tracking data, published daily
DeepSeek, OpenAI, Meta, Alibaba and Z.AI slashed input prices on eight models today, with cuts ranging from 13% to 80%. The biggest drop hit GLM 5.3 Flash from Z.AI, down 80% to $0.03 per 1M tokens.
The ecosystem gained 28,702 GitHub stars across 10,000+ tracked open-source AI projects today, and 66 new projects entered the index. This week 1,924 new projects and 100 new Hugging Face spaces were added.
The top gainer was dsh-Go, which added 3,112 stars. Two emerging projects were detected: awesome-dsh-plugin, a curated plugin list for DeepSeek Harness with 23,567 stars, and axuse, an accessibility-first browser tool for AI agents with 806 stars.
Releases of the day
Five minor releases shipped today. n8n 2.46.7 limits resume attempts for waiting and queued executions. Open Interpreter 0.0.60 added one canonical app server change with no model-provider catalog updates. CrewAI 1.15.39 documented the apps_not_connected error on /kickoff and /status endpoints. Morphic v1.10.0 fixed a draft handling issue and merged a panel feature. LoopX 2.3.0 introduced clearer acceptance and source handling for operators using managed Goals and Research Workspace.
Eight models saw input price reductions. DeepSeek V4 Pro 0423 dropped 63% to $0.21 per 1M tokens. OpenAI's gpt-oss-120b fell 33% to $0.02. Meta's Llama 3.1 8B Instruct dropped 60% to $0.02. Alibaba's Qwen3.6 27B fell 29% to $0.32. Z.AI cut GLM 5.2 batch by 44% to $0.39, GLM 5.3 Flash by 80% to $0.03, and GLM 5.3 batch by 13% to $0.39. Moonshot AI's Kimi K3 batch dropped 60% to $0.92.
▼ 63%DeepSeek V4 Pro 0423DeepSeek · input $0.56 → $0.21 per 1M tokens
▼ 33%gpt-oss-120bOpenAI · input $0.03 → $0.02 per 1M tokens
SplatAnything identifies a flaw in feed-forward 3DGS methods: Gaussian attributes are predicted in a radial-distance-normalized space that standardizes a global radial statistic of the reconstructed geometry, while the spacing remains unaddressed. The paper proposes Gaussian spacing normalization to fix this.
Today's broad price cuts across multiple providers signal continued downward pressure on inference costs, while the steady stream of minor releases and strong project growth suggest the ecosystem is expanding faster than any single model dominates.
Source: olud.ai tracking of 10,000+ open-source AI projects, 300+ models and live provider pricing. Figures are measured, not estimated. All releases · Live pricing · Latest in AI
