Ground Truth.
AI, checked against the source.
← 2026-09-212026-09-22later →

Xiaomi releases MiMo-V2.6, a trillion-parameter open agent family with a 9B distill

2026-09-22

Xiaomi released two one-million-context multimodal sparse agent models and a 9B Qwen distill, but its efficient Flash model still requires a roughly 178 GB weight download.

open-weights · agents · reinforcement-learning · mixture-of-experts · multimodal

Grok 4.7 improves agentic coding, but its gains consume more reasoning tokens

2026-09-22

xAI's Grok 4.7 offers a verified coding-agent improvement at the same standard API list price as 4.6, while independent testing finds it spends far more output tokens per task.

models · agents · coding · api-pricing · benchmarks

Linear rebuilt its CI after agent-assisted coding made validation the bottleneck

2026-09-22

Linear says its test suite nearly quadrupled as agents accelerated code production, and it cut critical-path overhead enough to hold pull-request waits just above five minutes.

coding-agents · developer-tools · ci · software-engineering · industry

Kev brings Jev-like local decision models to Qwen, while audits challenge universal calibration claims

2026-09-22

Kev released open local models for typed probability decisions, as published evaluations and an out-of-distribution audit show that confidence calibration varies sharply by task.

open-source · decision-models · calibration · local-ai · evaluation

Alibaba's verified Apsara news is a Qwen4 architecture preview and the Zhenwu M890 chip

2026-09-22

Alibaba has not publicly released Qwen4, but its Qwen3.8-Flash-Next repository previews the planned architecture and its T-Head unit has announced the 144 GB Zhenwu M890 accelerator.

open-weights · hardware · china · mixture-of-experts · inference

OpenAI says an internal model resolved 100-plus math problems and asks an independent group to advise on disclosure

2026-09-22

OpenAI has claimed more than 100 long-standing mathematical results from a new internal model, but has published neither an itemized list nor the proofs for that batch.

mathematics · scientific-ai · governance · formal-verification · research

Meta's Muse security design treats credentials, browser control, and network egress as separate agent boundaries

2026-09-22

Meta says Muse isolates credentials and mediates browser and network actions through separate services, while explicitly acknowledging that prompt injection remains an unsolved agent-security problem.

cybersecurity · ai-security · prompt-injection · agents · credentials · sandboxing

Apple's M5 Ultra makes large local AI practical by capacity, not by beating an RTX 5090

2026-09-22

A 256 GB M5 Ultra Mac Studio ran large sparse models and long contexts locally in a detailed review, but an RTX 5090 remained faster whenever the tested workload fit in its VRAM.

local-ai · hardware · apple-silicon · inference · open-weights

← 2026-09-212026-09-22later →