opencode regression: dual max_tokens fields break OpenAI-compatible endpoints
amonshiz · ghdev · 2026-10-15
anomalyco/opencode issue #55726 reports that since v2.0.28, requests to @ai-sdk/openai-compatible providers send both maxtokens and maxcompletiontokens, and strict endpoints (e.g. Baseten-hosted OpenAI-compatible APIs) reject any body containing both — verified via curl. GLM-5.3, GLM-5.3-Fast and Kimi-K3 are affected while OpenAI/Anthropic endpoints are not. The regression stems from #54694's migrateModel change; v1.19.0 and v2.0.27 work, v2.0.29 remains broken.
More from coding & agent
- Claude cut a 2.6-hour video export to 20 minutes — but only after being told to rebuild the server — byhewar · 2026-10-23
- 3 flight-booking rules my AI agent checks that I'd never do by hand — Efistoffeles · 2026-10-23
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Building an autonomous SEO agent in Hermes with three skills: programmatic pages, Keywords Everywhere, DataForSEO — EXM7777 · 2026-10-23
- Building a sports betting AI agent with onchain execution and human approval by default: architecture and lessons — Dry-Comfortable-2678 · 2026-10-23
- How do you audit AI agent decisions in regulated workflows? A field survey — Ok_Way_1993 · 2026-10-23