Unverified50% confidenceFactExact time
In Anthropic's Claude 3.7 and later series, the model's internal reasoning (thinking) and the final reply share the same max_tokens budget, and max_tokens is the total upper limit of 'thinking + output'.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedMoE models activate only a small fraction of total parameters per inference; Qwen3 Max may activate between 1/8 and 1/16 of the total parameters per inference.64% similarUnverifiedschemagate's reranker code calls provider.complete with max_tokens=120, causing the thinking process of Claude 5 series models to exhaust the entire token budget, and the model is truncated before outputting any ranking result.63% similarUnverifiedArgon's thinking_level parameter has three tiers: low, medium (default), and high; reasoning tokens are billed at the output price, even if the reasoning process is not shown to the user.59% similarUnverifiedClaude Code's recovery limit for max_tokens errors is three attempts, with the constant named MAX_OUTPUT_TOKENS_RECOVERY_LIMIT = 3.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1049511API
curl https://kongchang.com/api/v1/knowledge/claims/1049511MCP
get_claim(id=1049511)