Superpower Daily
XDOF Enters Series B Talks at a Reported $1.2 Billion ValuationArtificial Analysis Clarifies Astra’s 67-Point Coding Score Is...Anthropic’s Claude Built a 13-Million-Line Fermat Proof, Leavin...Microsoft Opens MAI-Image-2.6 in Foundry Preview and Adds a Low...Nvidia Agrees to Buy Hugging Face for $12.93B, Promising Hardwa...Databricks Says Proteus Generated Qwen Kernels Up to 5.2× Faste...XDOF Enters Series B Talks at a Reported $1.2 Billion ValuationArtificial Analysis Clarifies Astra’s 67-Point Coding Score Is...Anthropic’s Claude Built a 13-Million-Line Fermat Proof, Leavin...Microsoft Opens MAI-Image-2.6 in Foundry Preview and Adds a Low...Nvidia Agrees to Buy Hugging Face for $12.93B, Promising Hardwa...Databricks Says Proteus Generated Qwen Kernels Up to 5.2× Faste...

Full archive

Benchmark: Latest AI News

Search reporting on AI markets, models, products, companies, policy, and culture.

Category

More filters
BenchmarkClear all
Showing 1-10 of 10Results matching "Lean"

Today

1 story
BenchmarkBusiness

Artificial Analysis Clarifies Astra’s 67-Point Coding Score Is a Codex System Result

The updated methodology makes the score a clearer measure of a deployable coding agent. It also leaves a central question unanswered: how much performance comes from Astra itself versus the software layer around it. Artificial Analysis’ current coding-agent methodology settles the unit behind Astra’s 67-point result: it is GPT-6 Astra operating through Codex, not a score for the model in isolation.

5 min readRead story

Earlier this week

1 story
BenchmarkModels

Anthropic’s Fable 5.1 Takes Benchmark Lead, but Its Top Setting Costs More Per Task

Artificial Analysis found Anthropic’s newest general model reached its highest measured intelligence score, while the token use required at maximum effort raised the cost of its own evaluation tasks. Anthropic ’s Claude Fable 5.1 has taken the top measured position on Artificial Analysis’ Intelligence Index, scoring 66 at its maximum effort setting. Related on X Claude Fable 5.1 is now available in Cursor!

4 min readRead story

Last week

5 stories
BenchmarkModels

Hack The Box Benchmark: Best Human Team Cleared 36 Challenges; Top AI Team Reached 32

The latest benchmark points to a useful division of labor in cyber work: agents can accelerate solving, but elite performance still depends on people choosing paths and checking results. The strongest AI-augmented teams in Hack The Box’s benchmark worked three to four times faster than their human-only peers. Yet the only team to clear all 36 challenges was human; the best AI-augmented team stopped at 32.

3 min readRead story
BenchmarkModels

Chatbots Debunked Foreign Falsehoods About 75% of the Time, Beating Search Results

The result makes chatbots a potentially stronger starting point for investigating state-spread claims than a search-results page. It does not make their answers self-validating: language, citations and product design still shape the outcome. For someone confronting a false claim tied to a foreign influence campaign, a chatbot may be more likely to challenge the premise than a conventional page of search links.

4 min readRead story
BenchmarkProducts

Beijing Humanoid Games Put Practical Work to the Test; Only 3 of 12 Finish Fire Course

The games paired spectacle with multi-step chores and emergency work, where task switching, human control and offboard computing remain central constraints. The harder question at Beijing’s World Humanoid Robot Games was not how fast a robot could run, but whether it could complete a changing job. Human teleoperation was allowed in some events, with score penalties for remotely controlled robots.

2 min readRead story
BenchmarkProducts

Tiangong Ultra Runs 100m in 8.86 Seconds, Then Hits a Stopping Mat

The Beijing race shows a sharp jump in humanoid sprint performance, but its finish also exposed a basic operational constraint: Tiangong Ultra needed a thick mat to stop and caught fire after impact. Tiangong Ultra, a humanoid robot from the Beijing Humanoid Robot Innovation Center, ran 100 metres in 8.86 seconds in a large-size division semifinal at Beijing’s World Humanoid Robot Games. High jump: 2.88 metres.

3 min readRead story

August 2026

3 stories
BenchmarkTools

InferenceX Adds 11 Telemetry Views to AgentX, Exposing What Benchmark Curves Hide

The new exploration layer makes cache setup, bursty subagents and queue behavior inspectable, but its best-curve design means readers must still examine each point’s underlying configuration before treating a curve as a like-for-like comparison. A throughput curve can make two inference systems look nearly identical. The distinction reflects the workload AgentX is designed to measure.

3 min readRead story
BenchmarkModels

Inherent Says Its 27B Science Agent Beat OpenAI and Anthropic on Paper Replication

The company’s result centers on reproducing known findings, not making discoveries. Its next test is whether that training approach can move from replication to reliable new science. Inherent says its Faraday agent independently reproduced findings from published scientific papers without receiving the answers in advance, outperforming Anthropic ’s Claude Opus 4.8 and OpenAI ’s GPT-5.5 in the company’s evaluation.

2 min readRead story
BenchmarkInfrastructure

Nvidia Dynamo’s Agentic Speedups Target the Work Around the Model

The reported gains come from changing how serving systems route, retain and stream state under concurrency. The next challenge is whether these AgentX-specific improvements hold across production mixes and unfinished scheduling work. NVIDIA Dynamo’s agentic-serving optimization work treats the router, request plane and cache-ownership ledger as performance-critical infrastructure—not background plumbing.

4 min readRead story