Fable 5.1 Tops Benchmarks but Running Costs Climb

After Anthropic released Claude Fable 5.1, a sharp divergence between performance and cost sparked widespread discussion. Artificial Analysis's in-depth evaluation shows Fable 5.1 (max) scored 66 on the Intelligence Index, surpassing Opus 5 and GPT-5.6 Sol to take the top spot, but costs $3.76 per task—20% more than Fable 5 (max)'s $3.14—while total running costs are 56% higher than the previous generation. Multiple users independently reported and verified this cost increase. TL;DR: Fable 5.1 does top the charts, but at the default high-effort setting costs went up rather than down—whether it's worth it depends on how you use it.

Confirmed

• Artificial Analysis testing: Fable 5.1 (max) scored 66 on the Intelligence Index, beating Opus 5 and GPT-5.6 Sol; output tokens vary 11x across the five effort settings (about 13.1M at low, up to 143.7M at max)
• Cost breakdown: $3.76 per Intelligence Index task, 20% more than Fable 5 (max)'s $3.14; cache price cuts saved Anthropic $1.40, but max's roughly 1.7x output tokens pushed per-task cost above the previous generation
• Token consumption on the AAII task surged from about 83 million on Fable 5 to 144 million, a 73.5% increase
• Cache read prices dropped 75%; Anthropic says typical scenarios can save 25% on costs
• User @WonderFactory measured $3.69 per task in real testing, higher than Fable 5
• @haider1 cited CursorBench 3.2 data: Fable 5.1 Medium scored 68.0% at $3.53 per task, beating GPT-5.6 Sol Max's 67.2% at higher cost
• Fable 5.1 topped the original on Senior SWE-Bench with better pass^3 tiebreak performance, with Medium as the best setting

User reports and controversy

• @tetsuoai reported a single prompt burning through 20x usage quota in 52 minutes with no code returned, warning that the model spawns multiple Agents and recommending limiting it to a single Agent to control costs
• @MartinGTobias relayed user complaints of burning through half a month's quota in one day
• @op7418 summarized that beyond the 20% cost increase, there are issues like no support for forced tool choice, older models unable to read thinking, and mentioned invisible watermarks
• @VraserX observed Pro quota draining faster than Opus 5, speculating either extremely high compute demands or significantly improved capabilities
• Some users accused Anthropic of continuously raising prices, contrasting with OpenAI's price-cutting trend; @Angaisb responded that the high-end version actually costs less than Fable 5 while scoring higher

Why it matters

• Value for money depends on setting choice: a low-cost mode shared by an Anthropic engineer shows low-effort matching high-effort on CursorBench at just 1/3 the cost; @danielmac8 noted Fable 5.1 Low beats Mythos 5 High on Terminal-Bench 4.0
• @scaling01 offered a "cost per unit of intelligence keeps falling" perspective: despite rising absolute costs, the spend for equivalent intelligence is dropping; the same author also cited the 56% total cost increase, showing both narratives coexist
• The episode highlights the token-explosion problem of reasoning models at high-effort settings, posing a real challenge for subscriber quota management

2026-09-02 ~ 2026-09-02 · 32 related posts

Full story(7 episodes)→

• Episode 1: Claude Fable 5.1 Spotted on Bedrock and in Docs Ahead of Imminent Launch(2026-09-01, 5 posts)
• Episode 2: Anthropic Releases Claude Fable 5.1 and Mythos 5.1 with Major Agent Upgrades and 75% Cache Price Cut(2026-09-02, 91 posts)
• Episode 3: Fable 5.1 Tops Benchmarks but Running Costs Climb(2026-09-02, 32 posts)
• Episode 4: Every's Vibe Check: Anthropic Fable 5.1 Is So Back, With Half the Token Cost(2026-09-02, 9 posts)
• Episode 5: Anthropic's Fable 5.1 Tops Multiple Benchmarks as Cost Concerns Emerge(2026-09-02, 8 posts)
• Episode 6: Fable 5.1 Fixes Stiff 'Claude-Speak', Users Report More Coherent Output(2026-09-02, 6 posts)
• Episode 7: Fable 5.1 Rate Limits and Suspected Claude Code Caching Bug Spark User Backlash(2026-09-02, 6 posts)

• Claude Fable 5.1 Launches: High Usage Hints at Major Compute Jump — VraserX · 2026-09-02
• Tips for Claude Fable 5.1: Low-Effort Mode and Cost Optimization — RLanceMartin · 2026-09-02
• Fable 5.1 Beats GPT-5.6 on Benchmark at Lower Cost — haider1 · 2026-09-02
• Hands-on review: Claude Fable 5.1 matches Opus at lower cost with prompt cache savings — every · 2026-09-02
• Fable 5.1 leaps ahead on price-performance, beating Sol 5.6 Max on Cursor Bench for less — kimmonismus · 2026-09-02
• Fable 5.1: Matches High-End Rivals at Lower Cost — daniel_mac8 · 2026-09-02
• Fable 5.1 Cost Found Higher Than Fable 5 in Real-World Test — WonderFactory · 2026-09-02
• Fable 5.1 is 56% more expensive to run than Fable 5 — Angaisb_ · 2026-09-02
• Fable 5.1 token usage surges 73.5% on AAII benchmark — Angaisb_ · 2026-09-02
• Fable 5.1 Max Reasoning Costs 56% More Than Version 5 — Angaisb_ · 2026-09-02
• Users Question Anthropic's Pricing Strategy vs OpenAI's Deflation — Angaisb_ · 2026-09-02
• Fable 5.1 Matches Performance of Fable 5 at Half the Cost — scaling01 · 2026-09-02
• Fable 5.1 Benchmarks Show Gains Over Sol 5.6 Max — eyishazyer · 2026-09-02
• [source] Claude Fable 5.1 tops benchmarks but costs 20% more per task due to higher token usage — ArtificialAnlys · 2026-09-02
• [source] Fable 5.1 costs $3.76 per task; Anthropic's cache price cut saves $1.40 — ArtificialAnlys · 2026-09-02
• [source] Claude Fable 5.1's five effort levels span 11x in token usage, 58-66 on AA Index — ArtificialAnlys · 2026-09-02
• Fable 5.1 costs 6x more than GPT-5.6 Sol in Three.js test — rohanpaul_ai · 2026-09-02
• Fable 5.1 Costs 6x More Than GPT-5.6 in Three.js Generation Test — rohanpaul_ai · 2026-09-02
• Fable 5.1 tops SWE-Bench, matching prior perf at 50% cost — ajratner · 2026-09-02
• Fable 5.1 drains Claude 5-hour limit in under 30 minutes, users report — robleclerc · 2026-09-02
• New Claude Model Burns Through Usage: 20x Limit in 52 Minutes via Single Prompt — tetsuoai · 2026-09-02
• Fable 5.1 Hike: User Reports Half Monthly Credits Consumed in One Day — MartinGTobias · 2026-09-02
• Fable 5.1 review: Burns through 20x usage in 52 minutes on single prompt — tetsuoai · 2026-09-02
• Fable 5.1 'pure insanity': user says it one-shotted his monthly usage limit — talkaboutdesign · 2026-09-02
• Claude 5.1 Costs Rise 20% Despite Cache Cuts Due to Bloat and Watermarks — op7418 · 2026-09-02
• Fable 5.1 slashes cache prices by 75%, cutting total Agent costs by up to 45% — oran_ge · 2026-09-02
• Fable 5.1 launches on Cola: list price drops 30%, but Max mode costs rise — oran_ge · 2026-09-02
• Cache Price Cut May Not Lower Costs: Fable 5.1 Outputs More Tokens — eyishazyer · 2026-09-02
• Fable 5.1 burns through Max users' 5-hour limits in minutes; here's what's really going on — APPSO · 2026-09-02
• Anthropic's Fable 5.1 claimed 45% savings, but Max users burn limits in under an hour — heypearlai · 2026-09-02

2 near-duplicate retellings: scaling01 · Angaisb_