Unverified50% confidenceFactExact time
agentgateway's Ollama provider goes through the OpenAI-compatible endpoint, which completely ignores Ollama's native "think": false parameter, causing reasoning models' chain of thought to consume the entire max_tokens budget, the content to return an empty string, finish_reason to be length, and the HTTP status code to still be 200.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
UnverifiedOllama uses llama.cpp as its inference engine at the underlying level and exposes a REST API compatible with the OpenAI specification (the /v1 endpoint).74% similarUnverifiedOllama is an open-source "headless" LLM server with no graphical user interface that provides model inference capabilities externally through a REST API.72% similarUnverifiedThe OpenAI-compatible endpoint is a lowest-common-denominator abstraction that unifies fields such as messages, max_tokens, and finish_reason, but cannot express vendors' proprietary capabilities (such as Ollama's think or Anthropic's extended thinking).72% similarUnverifiedOllama is a CLI, a model library, and an HTTP API running on port 11434, standing on top of an inference engine it did not write itself.68% similarUnverifiedOpen-source frontends and development tools such as Open WebUI, Continue, and Dify use Ollama as the preferred local inference runtime.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1047897API
curl https://kongchang.com/api/v1/knowledge/claims/1047897MCP
get_claim(id=1047897)