Monitoring of the z-ai/glm-5.2 endpoint at gmicloud/fp8 was retired as of 2026-09-04 because a single query costs more than the per-query guard allows, according to the tracking dashboard. The endpoin…
RouteWise, a multi-provider LLM routing system presented by Haoran Ni of Nanjing University, reduces cost by 26.6%, mean time to first token by 58.4%, and SLO violations by 98.2% compared with OpenRou…
Z.ai released GLM-5.3-Flash, a natively multimodal 320B-A18B model with a 1M-token context window, under the MIT license, priced at $0.15 per 1M input tokens and $0.50 per 1M output tokens. The model …
OpenRouter reported that daily token usage for OpenAI's GPT-5.6 Luna grew 13.8x and GPT-5.6 Terra grew 5.6x after price cuts of 80% and 20% respectively, citing Jevons Paradox. The platform's data sho…
Tencent released Hy4 preview, a 770-billion-parameter open-source flagship model with 49 billion active parameters and a one-million-token context window, targeting long-horizon software engineering, …
OpenRouter retired and delisted monitoring of the deepseek/deepseek-v3.2 endpoint at alibaba/fp8 on 2026-09-04 after it dropped out of the B3IT selection for 30 days, according to the tracking page. T…
Substructure AI released Subs, a pre-1.0 cloud-native agent harness that runs an unprivileged agent loop with no system access, using MCP servers for tools and supporting local or client-server operat…
A developer known as TokenOps on AWS has released three open-source packages—sayay-guard, styrr-llm, and tinkuy-agent—that together form what they call an 'AI Harness,' a set of traditional middleware…
Meta released Muse Glimmer, a 29.6-billion-parameter open-weight AI model designed for multi-step agent tasks, on August 10 under the Apache 2.0 license, and it runs locally on a single consumer GPU w…
Beijing-based Moonshot AI released its Kimi K3 model with open weights, positioning China's open software approach as safer than the closed strategies of Silicon Valley rivals OpenAI and Anthropic. Th…
Open-source and free AI models are becoming nearly as capable as frontier American models, with local models like Ornith 1.5 9B running on consumer hardware and free API models from Nvidia and others …
Tencent Hunyuan released and open-sourced Hy4 preview on Aug. 28, a large language model with 770 billion total parameters, 49 billion activated parameters, and a context window exceeding 1 million to…
Revalvo, a new local-first LLM evaluation workbench, launched on Product Hunt today, enabling users to run prompts against multiple models in parallel with per-response latency, tokens, and cost track…
Z.ai's GLM 5.3 Flash scored 81.0% on the HTB-Challenger Benchmark, solving 15 of 16 Hack The Box challenges, second only to OpenAI's GPT-5.6 Sol at 87.2%. Despite requiring more steps (median 24.5 vs.…
Apple's upcoming M5 Ultra, with 512GB memory and 1.2TB/s bandwidth, is projected to cost $15,000–17,000 when it ships in October, making local AI inference for models like DeepSeek V4 Pro, Kimi K3, an…
Z.ai, the Chinese AI company formerly known as Zhipu AI, revealed that the mysterious Ox Alpha model was an anonymous test of its GLM-5.3-Flash open-weight model, which it ran on OpenRouter and OpenCo…
OpenRouter has recorded $1.81 in total tracking spend on the anthropic/claude-sonnet-4.6 @ claude-on-aws endpoint since 2026-08-18, across 45,628 queries, with B3IT onboarding accounting for $1.5988 a…
Monitoring of the deepseek/deepseek-v4-flash-0731 endpoint at ambient/fp4 was retired after it stopped yielding usable samples, with the B3IT tracking status listed as retired and stalled since 2026-0…
Small, low-cost AI models are now available at prices that fundamentally change the economics of AI applications, with GPT-5.6 Luna priced at $0.20 per million input tokens and $1.20 per million outpu…
Dark Bloom, a distributed AI inference network, pays Mac owners roughly $37 per month for a Mac Studio with an M5 Ultra chip and 96GB of memory, requiring a 48GB RAM minimum and using Stripe for payou…