Where Fable 5 sits on the Artificial Analysis Intelligence Index
Claude Fable 5 costs about $0.32 per average coding-agent request on our measured traffic, at list prices of $10 per million input and $50 per million output tokens. Anthropic reported 80.3% on SWE-Bench Pro at launch.
A newer version is out. Claude Fable 5.1 →
Run Claude Fable 5 in your agent from $19/month
See plans| Cost at list price | Fable 5 |
|---|---|
| Prompt under 100K tokens50% of requests, 87% of input cached | $0.16 |
| Prompt of 100K to 272K39% of requests, 88% of input cached | $0.39 |
| Prompt over 272K tokens11% of requests, 86% of input cached | $0.87 |
| Average request | $0.32 |
| 100 requests | $32.47 |
| Median session16 requests; a session ends after 10 idle minutes | $4.49 |
| A month of regular usemedian of 20 users active on 5+ days | $497 |
Each logged request priced at the vendor’s list rates for its own token counts, including cache reads and any higher rate for long prompts. Average output: 698 tokens per request.
| Artificial Analysis | Value | Rank |
|---|---|---|
| Intelligence Index10 evaluations, 0 to 100 | 50 | 8 of 48 |
| Terminal-Bench 4.0agentic coding in a terminal | 42.4% | 9 of 39 |
| Output speedmedian | 67 tokens/s | 33 of 44 |
| Time to first answer tokenincludes thinking time | 2.3 min | 36 of 44 |
Source: Artificial Analysis (artificialanalysis.ai). Measured variant: Claude Fable 5 (Max, Opus 4.8 Fallback), speed on Anthropic. Fetched October 13, 2026. Rank is among the models on this site, older versions included. Artificial Analysis runs Terminal-Bench 4.0 in its own setup, so its scores differ from the tbench.ai leaderboard.
Fable 5.1 replaced it at the same list price with cache reads at $0.25 instead of $1 per million, which cuts the cost of our measured requests by 27%. Claude Opus 5.5 costs a third as much per request ($0.11) and scores higher on Terminal-Bench 4.0.
Claude Fable 5 was the first Mythos-class model Anthropic made generally available, on June 9, 2026, positioned above Opus 4.8. At launch Anthropic reported 80.3% on SWE-Bench Pro, against 58.6% for GPT-5.5. Fable 5.1 replaced it in September 2026 at the same $10/$50 list price with cheaper cache reads, and Opus 5.5 now outscores both. Artificial Analysis scores Fable 5 at 50 on its Intelligence Index and 42.4% in its Terminal-Bench 4.0 run, against 58 and 59.6% for Opus 5.5. With cache reads at $1 per million, it is the most expensive Claude per request in our measurements ($0.32).
Sources: Artificial Analysis
Existing Fable 5 setups that have not yet moved to Fable 5.1 or Opus 5.5.
Best model for Claude Code
Best model for Cline
Best model for Aider
Best model for OpenCode
Best model for OpenClaw
Best model for Kilo CodeNot sure which Claude to run? Best Claude model in 2026: Opus vs Sonnet vs Haiku →
Setting it up? Claude Fable 5: model ID, providers and agent config →
The agent matters as much as the model. Which agent gets the most out of Fable 5? →
Every model on this page is included with Standard Compute
