Realtime AI News
Anthropic releases Opus 5.5 with lower prices and Fable-level performance
Anthropic has released Opus 5.5, describing it as the strongest-performing model the company has tested to date. The new flagship arrives with lower pricing and performance the report places at Fable level, putting cost alongside capability as a headline claim.
Anthropic has released Opus 5.5, a new flagship model that the company describes as the strongest-performing model it has tested to date. TechCrunch reported the launch on September 22, framing the update around two claims: lower prices and performance at Fable level.
The official characterization is the most loaded part of the announcement. Calling a model the strongest the company has tested is a direct capability claim, and it sets the bar against which the release will be judged once independent evaluations arrive.
Pricing is the second explicit change. According to the report, Opus 5.5 costs less than what came before, which places cost reduction alongside capability as a headline feature rather than an afterthought.
The naming suggests an incremental update rather than a generational jump. Positioning Opus 5.5 against Fable implies Anthropic is trying to bring a higher tier of performance down to a price point more teams can reach.
For enterprises and developers, the practical effect is a shift in the model-selection calculus. When a top-tier model gets cheaper while matching a stronger tier on capability, the threshold at which teams move critical workloads up to the flagship model drops.
What remains unverified is the detail behind the claims. The available summary does not include benchmark scores, context window, or distribution channels, so the Fable-level description is currently an Anthropic framing rather than a reproduced result.
What to watch next: whether third-party evaluations confirm the performance claim, and whether the lower price forces competitors to adjust their own flagship pricing in response.
Why it matters
If independent evaluations reproduce the performance claim, the lower price could lower the barrier to flagship-model adoption and push rivals to adjust their own pricing.
Nearby Updates
All09/22, 23:46
Intel says the agent era is pushing data center CPU-to-GPU ratios toward 1:1
At its Intel Connection conference in Suzhou, Intel argued that the CPU-to-GPU ratio in data centers is moving from 1:4 toward nearly 1:1 in the agent era, making CPU value worth re-measuring. Executives described data preprocessing, KV cache compression, high-performance storage and CPU/GPU/NPU coordination for scenarios including home AI, in-car AI and robots.
09/23, 02:00
OpenAI launches GPT-6 Sol and Luna, pitching lower cost and fewer mistakes
OpenAI has launched two new models, GPT-6 Sol and GPT-6 Luna, describing them as cut from the same cloth as Astra while pitching lower cost and fewer mistakes. TechCrunch reports the pair arrived together, a sign that price and reliability, not just benchmark scores, are now the headline claims of a frontier release.
09/22, 23:00
Five AI safety sessions every founder should have on their TechCrunch Disrupt 2026 agenda
Five AI safety sessions every founder should have on their TechCrunch Disrupt 2026 agenda. At TechCrunch Disrupt 2026, five sessions across the AI Stage and Real World AI Stage cover AI safety, featuring leaders from Anthropic, Nvidia, AWS, Waabi, and more. Register before September 25 to save up to $200.
09/22, 23:00
AstroForge puts AI in command of its next spacecraft with a small transformer model
TechCrunch reports that AstroForge's next spacecraft, Autonomy-1, will be commanded by a small, transformer-based AI model. The choice shifts mission control logic away from purely ground-issued commands and brings modern AI architectures into a deep-space setting.