A group of Berkeley CS students built autonomous AI agents that replace chat interfaces with a system using task planning, sandboxed Python execution, SQLite storage, and email aggregation. The agents…
A developer built a local-first RAG tool called Study Assistant that implements a multi-provider fallback chain to ensure reliability. The system prioritizes local Ollama inference but automatically s…
AgentHub has launched an open-source API gateway that provides unified access to over 100 AI models, including DeepSeek, Qwen, and Claude, at a flat rate of $0.14 per million tokens. This pricing is a…
A developer built PyCodeIt, a free Python coding challenge platform that generates unique problems using a Bring Your Own Key (BYOK) architecture with Groq and Supabase. The platform keeps server cost…
A developer compiled a list of free AI API keys and credits for agentic coding, including OpenRouter, AgentRouter, FreeModel, Aerolink, and Tokeness. The guide advises testing services one at a time d…
Xnovelist, a free and open-source novel-writing tool, launches as a local-first alternative to subscription-based AI writing services like Sudowrite and NovelAI. The app runs entirely in the user's br…
Nod, a new AI notepad for macOS, captures and transcribes meetings, calls, and voice notes without joining the call or storing audio, processing audio in five-second chunks and discarding it immediate…
Step 3.7 Flash, released on May 29, 2026, is a structural upgrade to 3.5 Flash with a new vision encoder, runtime escalation, and a compute-control flag. The migration requires two environment variabl…
Openfusion, an open-source drop-in compound-model proxy, lets users point any OpenAI-compatible tool at it to fan out prompts to a panel of LLMs in parallel, then a judge model synthesizes a single an…
NVIDIA released Nemotron 3 Ultra on June 4, 2026, a 550-billion-parameter open-weights model that achieves the highest intelligence score among US open models. The model uses a hybrid Mamba-Transforme…
GLM 5.2, a new open-weights model, achieved 15% fewer achievements than Gemini 3 Flash in text adventure games, a statistically significant difference. The benchmark, costing $5.1, controlled for game…
Gptbased launched a new LLM leaderboard that integrates LMArena rankings with live OpenRouter pricing, offering daily snapshots and email alerts when new models appear. The platform features eight cat…
TrustedRouter achieved a state-of-the-art score of 70.6 on the DRACO benchmark by fusing five models including open-weights models DeepSeek V4 Pro and Kimi K2.6, surpassing OpenRouter's best fusion of…
MiniMax released the M3 open-weight model, claiming it costs 5% of Claude Opus per task, achieves 59% on SWE-Bench Pro, and supports a 1-million-token context window at one-twentieth the compute of it…
Chinese AI lab Z.ai released GLM-5.2, a 753B-parameter open-weights text-only LLM with a 1 million token context window, under an MIT license. The model leads the Artificial Analysis Intelligence Inde…
Jacky Liang, Dev Rel Lead at OpenRouter, dropped 11 large language models into a 2D battle royale game for 30 matches. Grok 4.1 Fast won 13 games at $0.97 per win, while Claude Sonnet 4.6 won 5 games …
A Hacker News user asked the community which AI models power their products' daily features, revealing that many teams choose models based on cost and latency rather than rigorous evaluation. The user…
A developer launched Models Pie, a tool that ranks LLMs by user-defined tradeoffs between speed, cost, and quality using data from BenchLM, OpenRouter, and Artificial Analysis. Users adjust priorities…
FusionHarness, an open-source mixture-of-agents compound-model server, was released as a self-hostable alternative to OpenRouter's Fusion API. It fans out prompts to a panel of LLMs in parallel, uses …
A developer built a budget model comparing the real cost of Claude Code, OpenRouter, and image APIs, revealing that production costs often exceed listed prices due to retries, planning buffers, and fa…