OpenRouter, an AI model marketplace, shows that OpenAI's year-old GPT-OSS 120b model commands 36% of Anthropic's Opus 4.8 volume, illustrating market segmentation. The author replaced Gemma 4 26b with…
Retort, a framework for comparing coding stacks, reveals that a model's performance depends on the entire stack—language, weights format, serving engine, agent, context engine, sampling, and prompt—no…
A developer released Millwright, an open-source, self-hosted LLM router built in Rust, designed to sit between coding agents and GenAI workloads for cost savings and transparency. The tool supports pr…
A developer benchmarked time-to-first-token (TTFT) across AI gateways, finding LLM Gateway delivered the first content token 35% faster cold (906ms vs 1392ms) and 34% faster warm (814ms vs 1232ms) tha…
A benchmark comparing LLM Gateway and OpenRouter for Claude Haiku 4.5 found LLM Gateway is ~35% faster to first token cold (906 vs 1392ms) and ~34% faster warm (814 vs 1232ms) at the median. The TTFB …
Meta is reportedly developing an internal AI prompt routing platform to compete with OpenRouter, the inference aggregation startup valued at $1.3 billion after a $113 million Series B. The platform ai…
Octomind Cloud launched in beta as a persistent cloud compute platform for AI agents, solving three core problems: agent sessions lost when laptops close, API key sprawl across multiple providers, and…
Enterprises are increasingly pursuing 'AI sovereignty' by diversifying their AI supply chains with open-source alternatives, driven by rising costs and data sovereignty concerns, according to Lewis Li…
GMI Cloud, led by Alex Yeh, reported that signed customer commitments pushed contracted annual recurring revenue above $500 million as of July 21st, while live ARR grew 2.4x during the first half of 2…
TokenSwitch launches a routing layer that lets developers control which AI models their coding agents use, automatically selecting the cheapest capable model across approved providers like OpenRouter …
OpenBench v1, an open framework for measuring AI performance and efficiency for specific codebases and use cases, has been introduced as companies seek better ways to evaluate coding-agent harnesses. …
Agent swarms make local AI rigs cost-effective for the first time, according to developer testing. A single-agent session on a 2×3090 rig costs $0.80 in API-equivalent tokens, while a swarm of 32 agen…
Moonshot AI's Kimi K3, a 2.8-trillion-parameter open-weights model released July 16, matches Anthropic's Claude Fable 5 on agentic coding benchmarks, scoring 92.4% to Fable's 92.6% on software-enginee…
DeepSeek V3 0324, a 671B-parameter Mixture-of-Experts model with 37B active per token and a 128k context window, is available on DeepInfra at $0.24 per million input tokens, the cheapest listed provid…
DekaLLM offers the cheapest inference for Mistral Nemo 12B at $0.02 per million input tokens, delivering 3,444 points per dollar based on a general score of 62. The 12.2-billion-parameter model suppor…
Morph offers the cheapest API pricing for Qwen3.6 27B at $0.29 per million input tokens, according to live provider data refreshed about one hour ago via OpenRouter. The 27-billion-parameter model, re…
Anaconda has acquired Kilo Code to give enterprises visibility and control over AI agent usage, addressing the lack of oversight as developers run AI tools on personal accounts with sensitive data. Ki…
Plasma AI released Fractal, an open-source tool that uses recursive agent loops organized in a tree of git worktrees to autonomously decompose and complete complex, multi-step tasks. Each node iterate…
TRMNL has launched a public beta of an AI Agent that lets users build plugins by typing natural-language prompts, such as "build me a plugin showing populations around the world." The agent requires a…
A Reddit thread revealed that the easiest way to try Kimi K3 is via Moonshot's OpenAI-compatible API, not local inference. The discussion highlighted that users sought alternatives after Claude began …