Latest updates from the world of artificial intelligence
OpenAI's two cheaper GPT-6 models took on the three briefs GPT-6 Astra answered in September, and the flagship won every round. Astra finished with 10 points, Sol took 5 and Luna took 3, the same order as the price list. Sol and Luna built each brief two to six times faster, and every one of their six builds shipped a fault you can see.
BitsMinds Lab
Read more →OpenAI released GPT-6 Sol and GPT-6 Luna on 22 September at $2/$10 and $0.10/$0.50 per million tokens, half what their GPT-5.6 predecessors cost. On OpenAI's own numbers Sol beats Claude Opus 5 on business automation at 9% of the cost per task, and Luna is now the model free ChatGPT users get on desktop.
OpenAI
Read more →Claude Opus 5.5 beat Fable 5.1 and GPT-6 Astra on the three briefs this series gives every frontier model: 10 points to Fable’s 9 and Astra’s 8. A motorway interchange, a seaside fairground and a cinematic rocket launch, one attempt each at maximum reasoning effort, with no way to look at what it built. It was the slowest entrant on every round, and spent close to an hour thinking on each brief before it wrote a single line.
BitsMinds Lab
Read more →Anthropic shipped Claude Opus 5.5 on 22 September at $4 and $20 per million tokens, a fifth under Opus 5 and ahead of it on every benchmark it published. The sticker fell 20% but Anthropic claims a task costs 40% less, because the model also spends fewer tokens getting there. Every number here is Anthropic's own.
Anthropic
Read more →The same one-file game brief from September, handed to three models that had never attempted it: one from Anthropic, one from OpenAI, one from xAI. All three passed every requirement the brief actually stated. All three then missed something it did not. Play all three builds in the article.
BitsMinds Lab
Read more →SpaceXAI’s Grok 4.7 took the three briefs this series has handed every frontier model — a motorway interchange, a seaside fairground and a cinematic rocket launch — against Claude Fable 5.1 and GPT-6 Astra. One attempt each, maximum reasoning effort, no looking at what they built. Every mechanism Grok was asked for is present and computed correctly. In all three, the thing a viewer actually watches was never connected to it.
BitsMinds Lab
Read more →SpaceXAI released Grok 4.7 on 21 September at exactly the price Grok 4.6 carried — $2 per million input tokens and $6 per million output. It beats its predecessor on every benchmark the company published, trails Claude Fable 5.1 on most of them, and costs a fifth as much to feed and an eighth as much to read.
SpaceXAI
Read more →Anthropic shipped Claude Opus 5.5 on 22 September at $4 and $20 per million tokens — the price and the day the leak named. The sixteen days before it were a different story: one unconfirmed release, three different names, and a chain that had already been wrong about two of them.
Anthropic model documentation
Read more →The Shanghai lab has put a 600-billion-parameter sparse mixture-of-experts model behind a live API, with 27 billion parameters active per token, a one-million-token context window and open weights promised for 15 October. Artificial Analysis scores it 44 on its Intelligence Index, at $1.00 per million input tokens.
MarkTechPost
Read more →Alibaba shipped an omni-modal model that takes text, images, audio in 113 languages and two-hour video into one million-token context, at $0.15 per million input tokens — five times under Gemini 3.8 Flash. Its agentic perception reads long video selectively, gaining accuracy while cutting tokens 45.7%. The weights are not open.
MarkTechPost
Read more →Both models rebuilt the same racing game from one brief. One chose a muted desert, the other bright arcade colour. Both are playable here. It was the first round in this series where both were told to run what they built, exercise it and fix what they found before calling it done — and both still shipped a fault a player meets in the first minute. Fable 5.1 took it 15–14, the narrowest result this series has recorded.
BitsMinds Lab
Read more →Google’s new speech-to-speech pair takes the top of the Artificial Analysis quality index — and the cheaper sibling runs at about a seventh of GPT-Live-1 Astra’s hourly cost. Both reason in the background without stopping the conversation.
Shanghai AI Laboratory dropped a 744-billion-parameter, MIT-licensed agent model on GitHub with no announcement at all — then published a 143-author technical report three days later. It tops five of 16 benchmarks and trails Claude Opus 5 badly on code.
arXiv
Read more →Anthropic has announced no Fable 5.2, and no Opus 5.1 or 5.2 either: the docs list four current models and nothing newer. The rumour rests on a single X account whose first date already passed. But the company has shipped six frontier models since April at a median gap of 24 days, which puts the next one due right about now.
Anthropic model documentation
Read more →The Tokyo lab shipped two orchestrators on September 11: a cost-first Fugu Max at $2 per million input tokens, and a capability-first Ultra v2 that takes best or joint-best on five of eight benchmarks with Claude Fable 5, Fable 5.1 and GPT-6 Astra deliberately kept out of its pool.
Sakana AI
Read more →
The same three build briefs, one attempt each, no browser — but handed to the middle of each lab’s range instead of the top. Anthropic’s Claude Sonnet 5 and OpenAI’s GPT-5.6 Terra both broke the same fairground ride in the same new way, and one of them shipped a game that never runs. All six builds are live in the article.
BitsMinds Lab
Read more →
Claude Opus 5 and GPT-5.6 Sol got the same three build briefs — a playable Pac-Man, an animated seaside fairground and a motorway interchange — one attempt each, with no browser to check their own work. All six builds run live in the article: play both games, watch the rest, and see which one holds up.
BitsMinds Lab
Read more →OpenAI’s GPT-6 Astra and Anthropic’s Claude Fable 5.1 got the same three build briefs — a motorway interchange, a seaside fairground and a cinematic rocket launch — one attempt each, at maximum reasoning effort, and neither could open a browser to check its own work. All six builds run live inside the article, faults and all: a bridge that hides a truck mid-crossing, a carousel that is not doing what it looks like it is doing, and two rockets with very different ideas of what a launch is. Score it yourself.
BitsMinds Lab
Read more →
We gave four Claude models — Fable 5.1, Opus 5, Fable 5 and Sonnet 5 — the same five build briefs at maximum reasoning effort, one attempt each, with no browser to check their work. All twenty builds run live inside the article. Fable 5.1 won four of the five rounds and finished only a single point clear, because the round it lost was the one where its code never ran at all.
BitsMinds Lab
Read more →OpenAI's largest training run ever produced a model that drives a computer at superhuman speed and is the first rated Critical on cyber. Independent scoring puts its intelligence level with the model it replaces — at 2.5x the price.
Axios
Read more →Meta's fourth frontier model in five months takes the top spot on DeepSWE, near-saturates million-token retrieval, and undercuts every rival on price — as long as you let Meta train on your traffic.
The Register
Read more →Atlas is a single model that generates a minute of 1440p camera-controlled video, reconstructs scenes in 3D from a handful of photos, and outputs point clouds and Gaussian splats — beating specialist reconstruction models on their own benchmark.
World Labs
Read more →OpenAI says Astra is the first large language model to cross the Critical cybersecurity tier of its Preparedness Framework, scoring 100% on ExploitBench and discovering two zero-days on its own. Advanced cyber access goes to alpha testers first.
OpenAI
Read more →Anthropic’s Fable 5.1 doubles Terminal-Bench-Science, tops every published head-to-head against GPT-5.6 Sol, cuts cache reads 75% — and ships beside Mythos 5.1, a candidly riskier sibling for vetted defenders.
Anthropic
Read more →