Providers silently overwrote the deepseek-v4.1-flash model with -1017, reducing performance and increasing tool calls, according to 40 days of coding-agent telemetry
Read the original at www.reddit.com→Around mid-October deepseek-v4.1-flash suddenly started feeling much dumber in my coding agent, so I parsed my local session logs to see when it broke. Before the update it had a funny quirk where it would randomly...
Original headline: "Why you can't trust cloud API endpoints: 40 days of coding-agent telemetry showing the exact day providers silently overwrote deepseek-v4.1-flash with -1017 ("faster execution, shorter answers" = 3x more swearing and 6.7x more model switches)"