Expired50% confidenceBenchmarkExact time
In this 15-round evaluation, Claude Fable 5.1 took a total of 9 hours 35 minutes and cost $513.36; GPT-6 Astra took a total of 11 hours 19 minutes and cost $326.98 — Astra was approximately $186 cheaper overall but took about 1 hour 43 minutes longer than Fable.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
9/9/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 9/23/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedIn a Crossy Road-style game generation test, Fugu Ultra took 22 minutes, consumed approximately 90,000 tokens, and cost about $7.32, while Claude Opus 4 took 79 minutes, consumed approximately 940,000 tokens, and cost about $37.85.70% similarUnverifiedMainstream model inference costs are priced per million tokens: GPT-4o is approximately $5–15, Claude 3.5 Sonnet is approximately $3–15, and the total token count for multi-step agent tasks is often 10–50 times that of a single conversation.68% similarUnverifiedGPT-5.6 Sol achieved a record-high score of 73% on the DeepSWE benchmark Max tier at a cost of $22 per task, while the competing product Fable scored $839 on the corresponding tier.68% similarUnverifiedTaking GPT-4o as an example, the cost of a single 128K context call is approximately $0.3–$0.5, and a complete Deep Research task may involve 10–30 such calls, making the cost of a single research task as high as $5–$15.67% similarUnverifiedThe per-task cost of GPT-6 Astra is cheaper than that of the previous generation.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/881464API
curl https://kongchang.com/api/v1/knowledge/claims/881464MCP
get_claim(id=881464)