Vladimir Belomorski (Bello)

Three branches stripped bare.
Just the wind whispers to them;
Moon among the clouds..

[Vladimir Belomorski (Bello) @bellodox]

Three branches stripped bare.
Just the wind whispers to them;
Moon among the clouds..

[Vladimir Belomorski (Bello) @bellodox]

It shows exactly what the current problem with AI is: doing perfect job at the tests while performing subpar in real life tests. @claudeai #antropic

[Artificial Analysis @ArtificialAnlys]

Claude Opus 4.8 takes the lead on the Artificial Analysis Intelligence Index at 61.4, with Anthropic retaking the #1 spot on GDPval-AA and advancing in terminal use and scientific reasoning

To reach the leading position on the Intelligence Index, @Anthropic made large improvements in both real-world agentic work and frontier academic reasoning tasks.

Key takeaways:
➤ Claude Opus 4.8 is the new leader on the Artificial Analysis Intelligence Index. Opus 4.8 scores 61.4, up +4.1 points from Opus 4.7 and +1.2 points ahead of GPT-5.5 (xhigh), the previous Index leader

➤ The new release is slightly more efficient than its predecessor on agentic tasks, but token efficiency varied by task type. We saw Opus 4.8 use fewer turns and output tokens on GDPval-AA, but approximately the same number of output tokens for the overall Intelligence Index to achieve significantly higher performance.

➤ Anthropic retakes the lead on GDPval-AA, our primary evaluation for agentic performance on knowledge work tasks. Opus 4.8 scored an 1,890 Elo, reflecting an implied win rate of approximately 67% against GPT-5.5

➤ Claude is now among the top models for scientific reasoning. Previous releases have trailed peers on complex academic reasoning tasks, but with Opus 4.8, Claude sits slightly ahead of OpenAI and Google as the leader on Humanity’s Last Exam. It also scores higher than Gemini 3.1 Pro on CritPt, a frontier physics benchmark, but remains behind GPT-5.4 and GPT-5.5

➤ Claude Opus 4.8 reaches #2 on AA-Omniscience, slightly ahead of Opus 4.7. Opus 4.8 scores 27.4 on the AA-Omniscience Index behind only Gemini 3.1 Pro (32.9). Accuracy ticked up slightly to 46.6% and hallucination rate held roughly flat at 35.9% - Anthropic continues to demonstrate substantially lower hallucination rates than peer models from Google and OpenAI

➤ Compared with Opus 4.7, Opus 4.8 also makes material gains on Terminal-Bench Hard (+6.8 points), τ²-Bench Telecom (+5.9 points), and IFBench (+3.6 points), with relatively flat scores across AA-LCR, GPQA, and SciCode.

Other key model details remain the same as Opus 4.7:
Context window of 1 million tokens (equivalent to Opus 4.7)
Pricing of $5/$25 per million tokens of input/output; cache pricing remains at a 25% premium for cache writes ($6.25 per million tokens) with 5-minute time to live, and 90% discount for cache hits ($0.5 per million tokens)
Effort remains the recommended way of configuring model performance and latency, with the same options as Opus 4.7 - we measured the model at its ‘max’ effort setting to test peak performance

[Artificial Analysis @ArtificialAnlys]

Claude Opus 4.8 takes the lead on the Artificial Analysis Intelligence Index at 61.4, with Anthropic retaking the #1 spot on GDPval-AA and advancing in terminal use and scientific reasoning

To reach the leading position on the Intelligence Index, @Anthropic made large improvements in both real-world agentic work and frontier academic reasoning tasks.

Key takeaways:
➤ Claude Opus 4.8 is the new leader on the Artificial Analysis Intelligence Index. Opus 4.8 scores 61.4, up +4.1 points from Opus 4.7 and +1.2 points ahead of GPT-5.5 (xhigh), the previous Index leader

➤ The new release is slightly more efficient than its predecessor on agentic tasks, but token efficiency varied by task type. We saw Opus 4.8 use fewer turns and output tokens on GDPval-AA, but approximately the same number of output tokens for the overall Intelligence Index to achieve significantly higher performance.

➤ Anthropic retakes the lead on GDPval-AA, our primary evaluation for agentic performance on knowledge work tasks. Opus 4.8 scored an 1,890 Elo, reflecting an implied win rate of approximately 67% against GPT-5.5

➤ Claude is now among the top models for scientific reasoning. Previous releases have trailed peers on complex academic reasoning tasks, but with Opus 4.8, Claude sits slightly ahead of OpenAI and Google as the leader on Humanity’s Last Exam. It also scores higher than Gemini 3.1 Pro on CritPt, a frontier physics benchmark, but remains behind GPT-5.4 and GPT-5.5

➤ Claude Opus 4.8 reaches #2 on AA-Omniscience, slightly ahead of Opus 4.7. Opus 4.8 scores 27.4 on the AA-Omniscience Index behind only Gemini 3.1 Pro (32.9). Accuracy ticked up slightly to 46.6% and hallucination rate held roughly flat at 35.9% - Anthropic continues to demonstrate substantially lower hallucination rates than peer models from Google and OpenAI

➤ Compared with Opus 4.7, Opus 4.8 also makes material gains on Terminal-Bench Hard (+6.8 points), τ²-Bench Telecom (+5.9 points), and IFBench (+3.6 points), with relatively flat scores across AA-LCR, GPQA, and SciCode.

Other key model details remain the same as Opus 4.7:
Context window of 1 million tokens (equivalent to Opus 4.7)
Pricing of $5/$25 per million tokens of input/output; cache pricing remains at a 25% premium for cache writes ($6.25 per million tokens) with 5-minute time to live, and 90% discount for cache hits ($0.5 per million tokens)
Effort remains the recommended way of configuring model performance and latency, with the same options as Opus 4.7 - we measured the model at its ‘max’ effort setting to test peak performance

[Vladimir Belomorski (Bello) @bellodox]

Woo, the new Opus 4.8 behavior is quite erratic and inconsistent, @claudeai #aislop #antropic

[Vladimir Belomorski (Bello) @bellodox]

@godofprompt AI

[Vladimir Belomorski (Bello) @bellodox]

3 rules to express your thoughts so that everyone will understand you https://t.co/V1HG88ZMAb via @instapaper

[₵ryptospace ₮oday - All the news, analyzed daily @cryptospace2day]

Sneak Peek into the Future: A Tech Demo of SpaceXpanse Multiverse Features: https://t.co/NdAbEfFELi

[Vladimir Belomorski (Bello) @bellodox]

Building the metaverse of the future https://t.co/F7Wy2gEh7y via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

Compact nuclear fusion 1 million times more effective than other types, claims Israeli startup https://t.co/fF9dt1mglU via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

How freelancers in Austria can get a €100 health bonus in 2023 https://t.co/Q1pGRIWk88 via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

Is Vinegar an Acid or Base? And Does It Matter? https://t.co/XZoUuV0IVc via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

What's changing for tenants and homeowners in Austria in 2023? https://t.co/QUOgDvCWTN via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

5 Best Methods to Force Massive Muscle Growth | BOXROX https://t.co/hM099y0ipB via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

Най-добрите дни и часове за публикуване в социалните мрежи през 2023 г. (Инфографика) https://t.co/DWKpo3ziQF via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

EXPLAINED: How to pay Austria's TV and radio fee, or (legally) avoid it https://t.co/DhWuvyIfWD via @instapaper

[Vladimir Belomorski (Bello) @bellodox]

European Union to Put a 10,000-Euro Limit on Cash Payments; Transactions Over €1,000 in Crypto Will Be Scrutinized –… https://t.co/TFl6WI92LH

[Vladimir Belomorski (Bello) @bellodox]

Високоинтензивната физическа активност намалява риска от рак и метастази https://t.co/j7SklI1DP8 via @instapaper

[Bittrex Global @BittrexGlobal]

Day 12: It's the final day of #12DaysOfCrypto. We're giving away $3,000 worth of $MATIC to our lucky winners.
Enter now at: https://t.co/Tlat1xNJzg

🎁 Provide your KYC verified Bittrex Global email
🎁 Follow @BittrexGlobal, @OliverLinch and @MouradianMike on Twitter
🎁 Retweet!

[Bittrex Global @BittrexGlobal]

We're on day 11 of #12DaysOfCrypto! We're giving away $2,000 worth of $XRP to our lucky winners.
Enter now at: https://t.co/Tlat1xNJzg

🎁 Provide your KYC verified Bittrex Global email
🎁 Follow @BittrexGlobal, @OliverLinch and @MouradianMike on Twitter
🎁 Retweet!

[Vladimir Belomorski (Bello) @bellodox]

I just got my free Web3 Domain that matches my Twitter handle, bellodox.nft, - and you can, too! Follow the steps below.

1. Go here: https://t.co/RIezOx7lG4
2. Click "Verify Twitter"
3. Claim your domain

#web3 #ownyourname #UDfam