All Stories — Page 124
Every story published on RuntimeWire, newest first.
- Together AI adds Kimi.ai's K3 model, says it beats Claude Fable 5 on coding benchmark for about a third of the cost
The joint test of 452 DeepSWE rollouts shows K3 delivering near‑flagship coding quality at roughly 35% of Claude Fable 5's price, with higher pass@k scores.
- Premier League club Arsenal seeks AI engineer to turn AI research into coaching tools
The Premier League club announced a new role focused on turning its in-house AI research into actionable coaching and analysis software.
- World Foundation raises $52.5 million in WLD sale for identity push
Pantera led the WLD sale as Alex Blania tries to turn World's Orb-based network into proof-of-human infrastructure for companies and AI agents.
- Elon Musk says X will open‑source all code and undergo third‑party audit next month
In a July 24th, 2026 X post, Musk pledged total transparency for the platform by releasing its code and subjecting it to external review.
- Anthropic launches Claude Opus 5, says it matches Fable 5 intelligence for half the price
The new model, now on all paid Claude plans, hits state‑of‑the‑art scores on coding, ARC‑AGI‑3 and alignment tests, according to Anthropic.
- Runway Launches the First AI Router for Image, Video and Audio Models
The startup's new Runway Media Router, built into Runway Dev, lets developers route image, video and audio generation requests by quality, speed or cost.
- Foldkit packages Elm-style frontend architecture into a TypeScript framework
Eversound co-founder Devin Jameson's open-source project uses Effect to make state, side effects and debugging follow one prescribed loop.
- AMD's MI455X delivers 432GB HBM4, enough for trillion‑parameter model weights in one server
In a July 24th X post, AMD disclosed a new MI455X GPU that packs 432GB of HBM4, enough to store FP8 weights of a trillion‑parameter model in a single server.
- Ant Open Source releases LLaDA2.2-flash, an agent‑oriented MoE diffusion LLM with Levenshtein self‑editing
The open‑source community unveiled LLaDA2.2-flash on July 24th, combining mixture‑of‑experts diffusion with token‑level edit operations for on‑the‑fly correction.
- Patreon cuts 93 jobs as Jack Conte flattens management
Jack Conte says Patreon's growth remains steady, but the creator platform is cutting 93 jobs to lower costs while preserving its discovery push.
- Orca adds one-click resume for Codex and Claude Code sessions
The open-source agent platform lets developers continue a halted coding session when hitting usage caps on OpenAI's Codex or Anthropic's Claude Code.
- Okara rolls out AI-driven influencer agent to cut agency fees
The startup offers an end‑to‑end platform that finds creators, runs outreach and handles payments, aiming to replace $50k‑$100k agency launches.
- Jensen Huang uses first X post to back industry open-weight AI push
Nvidia's chief executive joined Satya Nadella and other industry leaders in signing an effort linking accessible model weights to US competitiveness, economic opportunity, and national security.
- Midjourney buys Co-Star as it builds its first standalone image app
The spring deal gives Midjourney an established mobile audience as David Holz expands beyond image models, Discord and the web.
- OrcaRouter launches free BYOK routing for developer-owned model keys
The routing option mixes developer-owned and platform keys as OrcaRouter builds a subscription business without token markups.
- Mimic Robotics connects FLUX 3 to industrial robots at Audi
The Zurich startup added factory data and an action decoder to Black Forest Labs' multimodal backbone for Audi manipulation tests.
- Letta releases open-source Trajectory package to standardize coding-agent data
The new format cuts token usage by up to fivefold and lets agents learn across Claude Code, Codex and Letta Code sessions.
- Supabase ships one-click Grafana monitoring on every plan
The free integration provisions Grafana Cloud and a dashboard covering more than 200 database and infrastructure metrics.
- SANA-Video 2.0 launched, promising efficient high-quality video generation
Enze Xie announced the model on X, highlighting a hybrid attention backbone and a new training pipeline that cuts compute while preserving quality.
- Figure reportedly manufactured its 1,000th humanoid robot in San Jose
The unconfirmed milestone would nearly triple Figure's official April count as Brett Adcock pushes humanoid robotics into factory-scale production.
- Make the Dot discloses $2.25M SAFE financing that began in 2024
The New York software maker has seven investors in the offering, but its SEC filing dates the first sale to August 15th, 2024.
- Sakana AI releases Fugu-Ultra v1.1, the latest version of its Fugu model family
The update adds unspecified upgrades and follows strong user adoption of Sakana AI's language models.
- Velmeni announces FDA clearance for AI-assisted 3D dental scan software
Velmeni is extending from 2D X-rays into CBCT, betting a unified imaging, voice and claims platform can beat standalone tools.
- Monk adds AI phone calls to its accounts receivable agent
George Kurdin and Joe Zhou are moving collections beyond email while keeping invoice access and payment changes under tight controls.