Anthropic claims its new Claude Opus 5 delivers near-Fable 5 performance at half the token price
Quick Answer
Anthropic's Claude Opus 5 achieves near Claude Fable 5 performance at half the cost, priced at $5 per million input tokens.
Quick Take
It excels in coding tasks, scoring 43.3% on Frontier-Bench v0.1, surpassing Fable 5 and GPT-5.6 Sol. Opus 5 also features improved safety filters and a new Fast Mode, enhancing speed by 2.5x at double the price.
Key Points
- Opus 5 costs $5 per million input tokens, half the price of Fable 5.
- Achieves 43.3% on Frontier-Bench v0.1, outperforming Fable 5 and GPT-5.6 Sol.
- New Fast Mode increases speed by 2.5x but doubles the token cost.
- Improved safety filters trigger 85% less often than Fable 5's.
- Opus 5 excels in generating visual outputs and tool creation.
DeepSignal Analysis
What happened
Anthropic has launched Claude Opus 5, which reportedly achieves performance close to Claude Fable 5 while costing half as much per million input tokens. It excels in coding tasks, scoring 43.3% on Frontier-Bench v0.1, outperforming both Fable 5 and GPT-5.6 Sol. The model also introduces a Fast Mode that increases processing speed by 2.5 times but at double the cost.
Key evidence
- Claude Opus 5 is priced at $5 per million input tokens, compared to $10 for Claude Fable 5, making it a more cost-effective option.
- On the Frontier-Bench v0.1, Opus 5 scored 43.3% in agentic terminal coding, surpassing Fable 5's score of 33.7% and GPT-5.6 Sol's score of 34.4%.
- Opus 5's safety filters trigger 85% less frequently than those of Fable 5, addressing previous criticisms regarding safety interventions.
Why it matters
The introduction of Claude Opus 5 at a lower price point may intensify competition in the AI model market, particularly against GPT-5.6 Sol and other Chinese models. Its strong performance in coding tasks and improved safety measures could attract users seeking both efficiency and reliability. However, the model's performance at maximum effort settings raises questions about its overall efficiency compared to its predecessors.
Source Excerpt
Anthropic's new flagship model Claude Opus 5 posts top scores in coding and knowledge work at half of Fable 5's token rates. On -3, a benchmark for novel problem-solving, Opus 5 hits 30. 2 percent, nearly four times higher than GPT-5. 6 Sol.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.
