Kimi K2.7-Code cuts thinking tokens 30% — but practitioners say the benchmarks don't check out

Kimi K2.7-Code claims 30% fewer thinking tokens and a drop-in API swap path, but independent benchmarks show kernel ...
TMCnet

KushoAI Benchmark Finds AI Coding Tools Struggle With Complex API Bugs

This report follows KushoAI's earlier launch of APIEval-20, the industry's first open benchmark for evaluating AI agents on ...

MiniMax-M3 debuts, eclipsing GPT-5.5 and Gemini 3.1 Pro on key benchmark performance for just 5-10% of the cost

M3 demonstrates that the next phase of agent development will not just be driven by larger datasets, but by efficient ...
Morningstar

KushoAI Unveils APIEval-20 to Benchmark AI Agents in API Testing

SAN FRANCISCO, April 8, 2026 /PRNewswire/ -- KushoAI, an AI-native platform for API testing and software reliability, has introduced APIEval-20, an open benchmark designed to evaluate how effectively ...
Tech Times

MiniMax M3 Open-Weight Coding Model: Frontier Claims, Unverified Benchmarks

MiniMax M3 launched June 1, 2026 with a 1-million-token context window and company-reported SWE-Bench Pro scores that edge ...
Morningstar

Octen Sets New Global Benchmark for Search Infrastructure; Launches World's Fastest Web Search API for the Agentic Era

Following a record-breaking sweep of global embedding leaderboards, Octen debuts its proprietary distributed search engine, achieving 60ms latency and 1M+ QPS to power search in the age of AI The ...