Unverified50% confidenceFactExact time
Major vendors routinely deploy models with internal codenames or temporary names on the Arena for testing prior to official release, and companies including OpenAI and Anthropic have tested new models there under hidden codenames.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/17/2026
First Seen
Valid until: 11/15/2026
Sources
Gemini 3.5 Pro变身3.7 Flash？大模型命名迷局解析
Related Claims
UnverifiedOnce Anthropic releases a new model, OpenAI typically follows shortly after with a competing product.71% similarUnverifiedCompanies such as OpenAI offer a "pinned model versions" feature, allowing enterprises to lock to a specific version until they actively choose to upgrade.69% similarUnverifiedAnti-injection techniques being researched by companies such as OpenAI and Anthropic include sandboxing data content, training models to refuse to execute data-layer instructions, and using a supervisory model to inspect the output of the execution model.66% similarUnverifiedThe official OpenAI team has released signals that native support for third-party large model integration may be added in the future.66% similarUnverifiedOpenAI models use the Codex environment in Frontier Code testing, Gemini models use Gemini CLI, and open-source models use Mini Agent.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/764535API
curl https://kongchang.com/api/v1/knowledge/claims/764535MCP
get_claim(id=764535)