Unverified50% confidenceFactExact time
The flagship models of leading institutions such as OpenAI, Anthropic, and Google DeepMind remain closed-source.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/5/2026
First Seen
Valid until: 1/3/2027
Sources
Related Entities
Related Claims
UnverifiedRepresentatives of the closed-source camp include OpenAI (GPT series), Anthropic (Claude series), and Google DeepMind (Gemini series), which do not release model weights and provide services exclusively through APIs.76% similarUnverifiedMajor frontier labs including OpenAI, Anthropic, and Google DeepMind have all signed commitments to provide AISI with evaluation access before model releases.74% similarUnverifiedCompanies such as OpenAI, Anthropic, Google, and Meta continue to release increasingly powerful foundation models, while open-source models like Llama and Mistral have already substantially narrowed the capability gap with commercial models.74% similarVerifiedLeading institutions such as OpenAI, Anthropic, and Google DeepMind possess frontier models and are deeply involved in government consultation, drafting of industry standards, and definition of safety evaluation frameworks.73% similarUnverifiedVendors including OpenAI, Anthropic, Google, and Meta continue to release foundation large models with comparable capabilities, and the gaps on mainstream benchmarks such as MMLU, HumanEval, and MATH have narrowed significantly.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/976254API
curl https://kongchang.com/api/v1/knowledge/claims/976254MCP
get_claim(id=976254)