Gemini 4 Neon scores 61 at its highest thinking setting on the Artificial Analysis Intelligence Index v5.0.1, ahead of Claude Opus 5.5 at 59. On Dec 1, Carbon’s standard API rates double, making Neon 25% cheaper per token than Carbon.
blog.invidelabs.com/gemini-4…
34
NGINX patched two HTTP/2 vulnerabilities. One can allow remote code execution under certain conditions and affects releases dating back to 2015. The upstream fixes are in 1.30.7 stable and 1.31.9 mainline. Here’s which build to update to.
blog.invidelabs.com/nginx-ht…
50
GPT-6.2 Astra cuts API rates 20%. OpenAI says Sol is within 0.5 pts of Astra on FrontierCode at 1/5 the cost, but it’s 3 pts below 6.1 Sol on an independent coding index.
Both report blockers less often. Codex adds a prompt to fix that; the API doesn’t.
blog.invidelabs.com/gpt-6-2-…
26
Qwen4-Max cuts input and output rates 25% from Qwen3.8-Max.
In Qwen’s launch chart, it leads on DeepSWE but trails Opus 5.5, GPT-6.1 Sol, GPT-6 Astra and Kimi K3.1 on Terminal-Bench.
We break down the benchmarks, pricing and open-weight plans.
blog.invidelabs.com/qwen4-ma…
59
Two engineers make progress, but neither knows who decides the first release. When a customer changes the priority, only one hears about it.
Our guide shows how to fix that, plus hiring, onboarding, delegation, feedback and growth.
blog.invidelabs.com/build-st…
28
Google’s Gemini 4 Hydrogen scores 13.5% on SWE-sweep versus Claude Sonnet 5.5’s 3.1% in Google’s selected results. Sonnet leads on Terminal-Bench 4.1.
The tests used different setups and sources. We look at the benchmarks, API pricing and cost per task.
blog.invidelabs.com/gemini-4…
37
A year after coding-agent adoption, a 655-firm study estimated:
• PRs opened: +35.7%
• PRs merged: +26.0%
• Time from submission to merge: +56.1%
How to spot where your PRs stall and decide when adding reviewers will help.
blog.invidelabs.com/ai-agent…
23
Claude Fable 5.2 cuts input and output token prices by 20% from 5.1. But at max effort, it cost about twice as much per task on one independent index.
We look at effort settings, Mythos access, Claude plans and API migration changes.
blog.invidelabs.com/claude-f…
1
32
Can an LLM improve another model by editing its weights?
GPT-6 Astra's reported Qwen edit reduced reasoning loops. Stronger unedited baselines complicate the math gains.
How weight-space programming works, and what this experiment shows:
blog.invidelabs.com/llm-weig…
1
1
82
OpenAI kept Git. Anthropic replaced it internally.
Their agent workloads led to different designs. We compare Pando, Scion and newer hosting and review tools, including ways to improve your workflow without changing forges.
blog.invidelabs.com/source-c…
19










